Gender Across Languages

<DOCINFO
AUTHOR ""
TITLE "Gender Across Languages: The linguistic representation of women and men. Volume II"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
Gender Across Languages

Impact: Studies in language and society
impact publishes monographs, collective volumes, and text books on topics in
sociolinguistics. The scope of the series is broad, with special emphasis on areas
such as language planning and language policies; language conflict and language
death; language standards and language change; dialectology; diglossia;
discourse studies; language and social identity (gender, ethnicity, class,
ideology); and history and methods of sociolinguistics.
General editor
Annick De Houwer
University of Antwerp
Advisory board
Ulrich Ammon William Labov

Gerhard Mercator University University of Pennsylvania
Laurie Bauer Elizabeth Lanza
Victoria University of Wellington University of Oslo
Jan Blommaert Joseph Lo Bianco
Ghent University The Australian National University
Paul Drew Peter Nelde
University of York Catholic University Brussels
Anna Escobar Dennis Preston
University of Illinois at Urbana Michigan State University
Guus Extra Jeanine Treffers-Daller
Tilburg University University of the West of England
Margarita Hidalgo Vic Webb
San Diego State University University of Pretoria
Richard A. Hudson
University College London
Volume 10
Gender Across Languages: The linguistic representation of women and men
Volume II
Edited by Marlis Hellinger and Hadumod Bußmann
Gender Across Languages
The linguistic representation
of women and men
volume 2
Edited by
Marlis Hellinger
University of Frankfurt am Main
Hadumod Bußmann
University of Munich
John Benjamins Publishing Company

Amsterdam/Philadelphia
TM The paper used in this publication meets the minimum requirements of American
8
National Standard for Information Sciences – Permanence of Paper for Printed

Library Materials, ansi z39.48-1984.
Library of Congress Cataloging-in-Publication Data
Gender Across Languages: The linguistic representation of women and men. Volume II /
edited by Marlis Hellinger and Hadumod Bußmann.
p. cm. (Impact: Studies in language and society, issn 1385–7908 ; v. 10)
Includes bibliographical references and indexes.
1. Grammar, Comparative and general--Gender. I. Hellinger, Marlis. II. Bußmann,
Hadumod. III. Impact, studies in language and society ; 10.
P325.5.P78 M4 2002
401.43--dc21 00-040314
isbn 90 272 18420 (Eur.) / 1 58811 0842 (US) (Hb; alk. paper)
isbn 90 272 18439 (Eur.) / 1 58811 0850 (US) (Pb; alk. paper)
© 2002 – John Benjamins B.V.
No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any
other means, without written permission from the publisher.
John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands
John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa
<TARGET
< R/ RE E F"toc"
F DOCINFO
AUTHOR ""
"lan">
"pref">
"ack">
"loa">
"hel">
"chin">
"ett">
"cha">
"zha">
"ger">
"eng">
TITLE "Table of contents"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
Table of contents
Languages of volume I and III vii

Preface ix
Acknowledgments xi
List of abbreviations xiii
gender across languages

The linguistic representation of women and men 1
Marlis Hellinger and Hadumod Bußmann
chinese
Editors’ note 27
In Chinese, men and women are equal – or – women and men are
equal? 29
Charles Ettner
Gender-related use of sentence-final particles in Cantonese 57
Marjorie K. M. Chan
Reality and representation: Social control and gender relations in
Mandarin Chinese proverbs 73
Hong Zhang
dutch
Towards a more gender-fair usage in Netherlands Dutch 81
Marinel Gerritsen
finnish
The communication of gender in Finnish 109
Mila Engelberg
< /R/TREARGET
E FF
"hal">
"gun">
"mar">
"bul">
"nis">
"pha">
"awb">
"not">
"ni">
"si">
"toc">
vi Table of contents
hindi
“Unnatural” gender in Hindi 133
Kira Hall
icelandic
Masculine generics in current Icelandic 163
Anna Gunnarsdotter Grönberg
italian
Gender and female visibility in Italian 187
Gianna Marcato and Eva-Maria Thüne
norwegian
The representation of gender in Norwegian 219
Tove Bull and Toril Swan
spanish
Gender in Spanish: Tradition and innovation 251
Uwe Kjær Nissen
vietnamese
Gender in addressing and self-reference in Vietnamese:
Variation and change 281
Hoa Pham
welsh
The politics of language and gender in Wales 313
Gwenllian Awbery, Kathryn Jones and Delyth Morris
Notes on contributors 331

Name index 337
Subject index 343
<TARGET "lan" DOCINFO
AUTHOR ""
TITLE "Languages of Volume I and III"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
Languages of Volume I
Arabic
Atiqa Hachimi
Belizean Creole
Geneviève Escure
Eastern Maroon Creole
Bettina Migge
English
Marlis Hellinger
Janet Holmes
Anne Pauwels
Suzanne Romaine
Hebrew
Yishai Tobin
Indonesian
Esther Kuntjara
Romanian
Florence Maurice
Russian
Ursula Doleschal and Sonja Schmid
Turkish
Friederike Braun
</TARGET "lan">
Languages of Volume III
Czech
Svetla Čmejrková
Danish
Kirsten Gomard and Mette Kunøe
French
Elisabeth Burr
Elmar Schafroth
German
Hadumod Bußmann and Marlis Hellinger
Greek
Theodossia-Soula Pavlidou
Japanese
Sachiko Ide
Janet S. (Shibamoto) Smith
Oriya
Kalyanamalini Sahoo
Polish
Gabriela Koniuszaniec and Hanka Błaszkowska
Serbian
Elke Hentschel
Swahili
Rose Marie Beck
Swedish
Antje Hornscheidt
<TARGET "pref" DOCINFO
AUTHOR ""
TITLE "Preface"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
Preface
The series Gender across languages is an ongoing project with potential follow-
up publications. Our main goal has been to provide a comprehensive collec-
tion of in-depth descriptions of gender-related issues in languages with very
diverse structural foundations and socio-cultural backgrounds. The project is
designed to have an explicit contrastive orientation in that basically the same
issues are discussed for each language within the same terminological and
methodological framework. This framework, whose central notion is, of
course, the multidimensional concept of “gender”, is discussed in the intro-
ductory chapter of “Gender across languages – The linguistic representation of
women and men”. Care has been taken not to impose a narrow western
perspective on other languages.
This is the second of three volumes which comprise a total of thirty
languages: (Moroccan) Arabic, Belizean Creole, Chinese, Czech, Danish, Dutch,
Eastern Maroon Creole, English, Finnish, French, German, Greek, Hebrew,
Hindi, Icelandic, Indonesian, Italian, Japanese, Norwegian, Oriya, Polish,
Romanian, Russian, Serbian, Spanish, Swahili, Swedish, Turkish, Vietnamese,
and Welsh. All contributions were specifically written for this project, in close
collaboration with the editors over a period of three years. Unfortunately, a few
languages (Bulgarian, Hungarian, Korean, Portuguese, and one Native Ameri-
can language) dropped out of the project for various reasons. These languages
should be included in a potential future volume.
The basis on which particular languages should be brought together in one
volume has been a problematic one to define. Rather than categorizing languag-
es according to language family (areal, typological or historical), or according
to whether the language has or does not have grammatical gender, or using an
overall alphabetical ordering, we decided – in agreement with the publisher – that
each volume should contain a balanced selection of languages, so that each volume
will provide readers with sufficient material to illustrate the diversity and complex-
ity of linguistic representations of gender across languages. Thus, each volume
will contain both languages with grammatical gender as well as “genderless”
languages, and languages with different areal, typological and historical affiliations.
</TARGET "pref">
x Preface
“Gender across languages” is, of course, a selection, and no claims can be

made that the three volumes will cover all language groups adequately. Critics
will find it easy to identify those language areas or families that are under-
represented in the project. In particular, future work should consider the
immense number of African, Asian and Austronesian languages which have so
far received little or no attention from a gender perspective.
Though we are aware of the fact that most languages of the project have
developed a number of regional and social varieties, with different implications
for the representation and communication of “gender”, we supported authors
in their unanimous decision to concentrate on standard varieties (where these
exist). This decision is particularly well-founded for those languages for which
gender-related issues are being described here for the very first time. Only in the
case of English, which has developed major regional standards with consider-
able differences in usage, did we decide to make explicit reference to four
different varieties (British English, American English, New Zealand English and
Australian English).
We took care that each chapter did address most of the questions we had
formulated as original guidelines which, however, were not intended (nor
interpreted by authors) to impose our own expectations of how “gender” is
represented in a particular culture. Thus, chapters basically have the same
overall structure, with variation due to language-specific properties as well as
to the state of research on language and gender in the respective country. In
some cases, we encouraged authors to include some of their own empirical
research where this has implications for the analysis of “gender” in the
respective language.
Marlis Hellinger
Hadumod Bußmann
<TARGET "ack" DOCINFO
AUTHOR ""
TITLE "Acknowledgments"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
Acknowledgments
Assembling a work of this scope requires the collaboration of many. The editors
were the fortunate recipients of a large amount of support: financial, scholarly,
technical, and moral. Therefore, it is a pleasure to record our gratitude to all
those who have been generous with their time, attention and expertise in
providing advice, references, and data, or have supplied helpful comments on
particular languages. We wish to thank specifically
– all contributors for their active cooperation and patience with our ongoing
demands during the preparation period over four years. Particularly those
colleagues who sent in their papers early had to endure a long wait for the
publication;
– all anonymous native-speaker reviewers, mainly members of the Board, for
their thorough data check und thoughtful council;
– our colleague Dr. Friederike Braun of the University of Kiel, Germany, a
professional general linguist and, in addition, a specialist on linguistic
gender studies, for her meticulous reading, expert editorial assistance and
insightful comments;
– our colleague Dr. Kerstin Kazzazi of the University of Eichstätt, Germany,
a professional general linguist and native speaker of English, for thoroughly
checking on the use of English;
– the staff of the Linguistics Department of the Institute of English and
American Studies at the University of Frankfurt: Helena McKenzie, Dr.
Bettina Migge, Dr. Susanne Mühleisen, Raimund Schieß, MA, Stephanie
Schnurr, MA, and Eva Tripp, MA, for their assistance with the final prepa-
ration of the manuscripts;
– the Deutsche Forschungsgemeinschaft (DFG) for financial support over the
period of two years.
Finally we would like to thank the team of John Benjamins Publishing Company,
especially Cornelis H. J. Vaes (acquisition editor), and the General Editor of
IMPACT, Annick De Houwer, PhD. (University of Antwerp), who were both
</TARGET "ack">
xii Acknowledgments
demanding and encouraging editors, and from whose sympathetic enthusiasm and
efficient expertise the final processing of the text has profited tremendously.
Marlis Hellinger, Frankfurt am Main Winter 2001

Hadumod Bußmann, München
<TARGET "loa" DOCINFO
AUTHOR ""
TITLE "List of abbreviations"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
List of abbreviations
abl ablative iness inessive

acc accusative inf infinitive
adj adjective instr instrumental
adv adverb ipf imperfective
agr agreement loc locative
allat allative masc/m masculine
anim animate n noun
ant anterior neg negation
aor aorist neut/n neuter
asp aspect nom nominative
att attributive np noun phrase
aux auxiliary obj object
caus causative obl oblique
cl classifier part participle
com comitative partt partitive
comp comparative pass passive
cond conditional past past
conj conjunction pers personal pronoun
cop copula pf perfective
dat dative pl/pl plural
def definite pol polite
dem demonstrative poss possessive
der derivational affix prep preposition
det determiner pres present
dur durative prog progressive
elat elative pron pronoun
erg ergative prt particle
ess essive refl reflexive
evid evidential rel relative pronoun/affix
expl expletive sg/sg singular
fem/f feminine subj subject
fut future top topic
gen genitive trans transitive
ger gerund trns translative
hab habitual verb verb
illat illative voc vocative
imp imperative 1 first person
inan inanimate 2 second person
indef indefinite 3 third person
</TARGET "loa">
<LINK "hel1-n*">
<TARGET "hel1" DOCINFO
AUTHOR "Marlis Hellinger and Hadumod Bußmann"
TITLE "The linguistic representation of women and men"
SUBJECT "Impact 9"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
Gender across languages

The linguistic representation of women and men*
Marlis Hellinger
University of Frankfurt am Main, Germany
Hadumod Bußmann
University of Munich, Germany
1. Aims and scope of “Gender across languages”

2. Gender classes as a special case of noun classes
2.1 Classifier languages
2.2 Noun class languages
3. Categories of gender
3.1 Grammatical gender
3.2 Lexical gender
3.3 Referential gender
3.4 “False generics”: Generic masculines and male generics
3.5 Social gender
4. Gender-related structures
4.1 Word-formation
4.2 Agreement
4.3 Pronominalization
4.4 Coordination
5. Gender-related messages
5.1 Address terms
5.2 Idiomatic expressions and proverbs
5.3 Female and male discourse
6. Language change and language reform
7. Conclusion
Notes
References
2 Marlis Hellinger and Hadumod Bußmann
1. Aims and scope of “Gender across languages”
“Gender across languages” systematically investigates the linguistic representa-

tion of women and men in 30 languages of very different structural and socio-
cultural backgrounds. Fundamental to the project is the hypothesis that the
formal and functional manifestations of gender in the area of human reference
follow general, and perhaps universal principles in the world’s languages. We
will outline these principles and specify the theoretical and empirical founda-
tions on which statements about gendered structures in languages can be made.
A major concern of “Gender across languages” is with the structural
properties of the individual language:
– Does the language have grammatical gender, and – if so – what are the
consequences for agreement, coordination, pronominalization and word-
formation, and more specifically, for the linguistic representation of
women and men?
– In the absence of grammatical gender, what are possible ways of expressing
female-specific, male-specific or gender-indefinite personal reference?
– Can asymmetries be identified in the area of human reference which may
be interpreted as the result of the choice of the masculine/male as the
default gender?
– What is the empirical evidence for the claim that in neutral contexts
masculine/male expressions are perceived as generic and bias-free?
– Does the language contain idiomatic expressions, metaphors, proverbs and
the like which are indicative of gender-related socio-cultural hierarchies or
stereotypes?
In addition, the project will outline gender-related tendencies of variation and

change, and – where applicable – language reform, seeking to identify the ways
in which the structural/linguistic prerequisites interact with the respective
social, cultural and political conditions that determine the relationships
between women and men in a community.
“Gender across languages” will focus on personal nouns and pronouns,
which have emerged as a central issue in debates about language and gender. In
any language, personal nouns constitute a basic and culturally significant lexical
field. They are needed to communicate about the self and others, they are used
to identify people as individuals or members of various groups, and they may
transmit positive or negative attitudes. In addition, they contain schemata of,
e.g., occupational activities and (proto- or stereotypical) performers of such
Gender across languages 3
activities. On a psychological level, an appropriate use of personal nouns may

contribute towards the maintenance of an individual’s identity, while inappro-
priate use, for example identifying someone repeatedly (either by mistake or by
intention) by a false name, by using derogatory or discriminatory language, or
by not addressing someone at all, may cause irritation, anger or feelings of
inferiority. And since an individual’s sense of self includes an awareness of
being female or male, it is important to develop an understanding of the ways
in which gender is negotiated in a language. This understanding must, of
course, be based on adequate descriptions of the relevant structural and
functional properties of the respective language.
In communication, parameters like ethnicity, culture, social status, setting,
and discourse functions may in fact be as important as extra-linguistic gender,
and none of these parameters is represented in a language in any direct or
unambiguous way (cf. Bing & Bergvall 1996: 5). Only a multidimensional
theory of communication will be able to spell out the ways in which these
parameters interact with linguistic expressions. By interpreting linguistic
manifestations of gender as the discursive result of “doing gender” in specific
socio-cultural contexts, the analysis of gender across languages can contribute
to such a theory.
Structure-oriented gender research has focused primarily on formal,
semantic and historical issues. On a formal level, systems of gender and
nominal classification were analyzed, with an emphasis on the phonological
and morphological conditions of gender assignment and agreement (cf.
Section 4.2).1
From a semantic perspective, a major issue was the question as to whether
the classification of nouns in a language follows semantic principles rather than
being arbitrary.2 While gender assignment in the field of personal nouns is at
least partially non-arbitrary, the classification of inanimate nouns, e.g., words
denoting celestial bodies, varies across languages. Thus, the word for ‘sun’ is
grammatically feminine in German and Lithuanian, but masculine in Greek,
Latin and the Romance languages, and neuter gender in Old Indic, Old Iranic
and Russian. Correspondingly, metaphorical conceptualizations of the sun and
the moon as female or male deities, or as the stereotypical human couple, will
also show variation.
Nominal class membership may be determined by conceptual principles
according to which speakers categorize the objects of their universe. The
underlying principles may not be immediately comprehensible to outsiders to
a particular culture. For example, the words for female humans, water, fire and
fighting are all in one nominal category in Dyirbal, an Australian language (cf.
Dixon 1972). The assignment of, say, some birds’ names to the same category
can only be explained by recourse to mythological association.3 Finally, histori-
cal issues in the study of linguistic gender concerned the origin, change and loss
of gender categories.4
Corbett’s account of over 200 languages is a major source for any discussion
of gender as a formal category. However, since Corbett analyzes entire noun
class systems, while we concentrate on personal nouns and pronouns, “sexism
in language” (Corbett 1991: 3) is not one of his concerns. But Corbett does in
fact contribute to that debate in various ways, in particular, by introducing
richness and diversity to a field which has been dominated by the study of a few
Western languages.
2. Gender classes as a special case of noun classes
Considering the lack of terminological precision and consistency in the debate

about language and gender, the terms “gender class” and “gender language”
need to be defined more precisely and with a more explicit reference to the
wider framework of nominal classification. Of course, it must be noted that
not all languages possess a system of nominal classification. In the project,
Belizean Creole, Eastern Maroon Creole, English, Finnish and Turkish5
represent this group of languages. Other languages may divide their nominal
lexicon into groups or classes according to various criteria. Among the
languages which exhibit such nominal classification, classifier languages and
noun class languages (including languages with grammatical gender) consti-
tute the two major types.6
2.1 Classifier languages

A prototypical case of classifier systems are numeral classifiers. In languages
with such a system, a numeral (e.g. ‘three’) cannot be combined with a noun
(e.g. ‘book’) directly, but requires the additional use of a classifier. Classifiers
are separate words which often indicate the shape of the quantified object(s).
The resulting phrase of numeral, classifier, and noun could, for example, be
translated as ‘three flat-object book’ (cf. Greenberg 1972: 5). Numeral classifiers
are thus independent functional elements which specify the noun’s class
membership in certain contexts. In addition, the use of classifiers may be
indicative of stylistic variation.
In languages with (numeral) classifiers, nouns do not show agreement with

other word classes, although classifiers may perform discourse functions such
as reference-tracking, which in gender languages are achieved by agreement. On
average, classifier languages have from 50 to 100 classifiers (cf. Dixon 1982:215).7
Classifier systems are rather frequent in East Asian languages, and in “Gender
across languages” are represented by Chinese, Indonesian, Japanese, Oriya and
Vietnamese.
2.2 Noun class languages

While in numeral classifier systems the class membership of nouns is marked
only in restricted syntactic contexts (mainly in the area of quantification), class
membership in noun class languages triggers agreement on a range of elements
inside and outside the noun phrase. Noun class languages have a comparatively
small number of classes (hardly more than 20). These classes consistently
structure the entire nominal lexicon, i.e. each noun belongs to one of these
classes (there are exceptional cases of double or multiple class membership).
French, German, Swahili and many others are noun class languages, but we find
these languages also referred to as “gender languages”.8 In accordance with
Craig (1994), we will not use the terms “gender language” and “noun class
language” synonymously, but will define them as two different types of noun
class languages based on grammatical and semantic considerations. This
distinction is also motivated by our interest in the linguistic representation of
the categories “female” and “male”.
“Gender languages”
This type is illustrated by many Indo-European languages, but also Semitic
languages. These languages have only a very small number of “gender classes”,
typically two or three. Nouns do not necessarily carry markers of class member-
ship, but, of course, there is (obligatory) agreement with other word classes,
both inside and outside the noun phrase. Most importantly – for our distinc-
tion – class membership is anything but arbitrary in the field of animate/
personal reference.
For a large number of personal nouns there is a correspondence between
the “feminine” and the “masculine” gender class and the lexical specification of
a noun as female-specific or male-specific. Languages of this type will be called
“gender languages” or “languages with grammatical gender”.9 The majority of
languages included in the project belong to this group: Arabic, Czech, Danish,
Dutch, French, German, Greek, Hebrew, Hindi, Icelandic, Italian, Norwegian,

Polish, Romanian, Russian, Serbian/Croatian/Bosnian, Spanish, Swedish, and
Welsh. As the examples of Oriya and English show, a gender system of this type
can erode (Oriya) and eventually be lost (English); cf. also Section 3.1.
“Noun class languages”

This type displays no obvious correspondence between class membership and
a noun’s specification as female-specific or male-specific in the field of personal
nouns. These languages, represented in the project by Swahili,10 have a larger
number of classes than gender languages. Often class membership is explicitly
marked on the noun itself (cf. the class prefixes in Bantu languages), and there
is extensive agreement on other word classes.
To summarize, we will speak of a “gender language” when there are just two or
three gender classes, with considerable correspondence between the class
membership and lexical/referential gender in the field of animate/personal
nouns. Languages with grammatical gender represent only one type of nomi-
nal classification requiring the interaction of at least two elements, i.e. of the
noun itself and some satellite element that expresses the class to which the
noun belongs.
The lack of grammatical gender in a language does not mean that “gender”
in the broader sense cannot be communicated. There are various other catego-
ries of gender, e.g., “lexical” and “social” gender, which may be employed to
transmit gendered messages. Thus, “gender languages”, languages with classifi-
ers or noun classes, as well as those languages that lack noun classification
completely (English, Finnish, Turkish), can resort to a variety of linguistic
means to construct gender-related messages.
3. Categories of gender
Having established the difference between the more comprehensive concept of

“noun class language” and the concept of “gender language”, it is necessary to
introduce a number of terminological distinctions beyond the typological level
which will focus more directly on the representation of women and men in a
language: grammatical gender, lexical gender, referential gender and social gender.

A central issue in any cross-linguistic analysis of gender is, of course, the
category of grammatical gender. Typically, gender languages as defined in
Section 2.2 have two or three gender classes – among them frequently “femi-
nine” and “masculine”. Sometimes the emergence of new subclasses presents
problems of analysis, examples being Serbian/Croatian/Bosnian and Russian
(cf. Corbett 1991: 161–168). By contrast, a language may reduce the number of
its grammatical gender classes, as in the case of some Germanic, Romance, and
most Iranian languages, or lose its original gender system completely, as
happened in English and Persian.11
Unlike case or number, grammatical gender is an inherent property of the
noun which controls agreement between the noun (the controller) and some
(gender-variable) satellite element (the target) which may be an article, adjec-
tive, pronoun, verb, numeral or preposition. Nominal gender typically has only
one value, which is determined by an interaction of formal and semantic
assignment rules.
3.2 Lexical gender

In debates on language and gender, the term “gender” usually relates to the
property of extra-linguistic (i.e. “natural” or “biological”) femaleness or
maleness. Thus, in English, personal nouns such as mother, sister, son and boy
are lexically specified as carrying the semantic property [female] or [male]
respectively, which may in turn relate to the extra-linguistic category of
referential gender (or “sex of referent”). Such nouns may be described as
“gender-specific” (female-specific or male-specific), in contrast to nouns such
as citizen, patient or individual, which are considered to be “gender-indefinite”
or “gender-neutral”. Typically, gender-specific terms require the choice of
semantically corresponding satellite forms, e.g., the English anaphoric pronouns
she or he, while in the case of gender-indefinite nouns, pronominal choice may
be determined by reference (e.g., to a known individual), tradition (choice of
“false generics”; cf. Section 3.4) or speaker attitude (as evident, e.g., from a
positive evaluation of “gender-fair” language). In languages with grammatical
gender, a considerable correspondence can be observed between a noun’s
grammatical gender class and its lexical specification, most consistently in the
field of kinship terms: Germ. Tante (f) ‘aunt’ and Onkel (m) ‘uncle’ have a
lexical specification as [female] and [male], respectively. Such nouns require the
use of the corresponding pronouns sie (f) and er (m). For terms without lexical
gender, i.e. gender-indefinite nouns such as Individuum (n) ‘indivdual’ or
Person (f) ‘person’, pronominal choice is usually, but not always, determined by
the grammatical gender of the antecedent (see Bußmann & Hellinger, vol. III).
We do not wish to imply that the terms “female-specific” and “male-
specific” correspond to a binary objectivist view that categorizes people neatly
into females and males. For example, anthropologists have discussed the
Hindi-speaking hijras as a “third gender”: “[…] most hijras were raised as boys
before taking up residence in one of India’s many hijra communities and
adopting the feminine dress, speech, and mannerisms associated with member-
ship” (Hall, this vol.).12 Although the terms “female” and ”male” contribute to
the construction of people’s everyday experience, they might perhaps be more
adequately placed on a continuum, which allows for variation, fuzzy category
boundaries, and prototype effects (cf. Lakoff 1987). In spite of this insight, we
will continue to use the terms “female” and “male” as valuable descriptive tools.
In any language, lexical gender is an important parameter in the structure
of kinship terminologies, address terms, and a number of basic, i.e. frequently
used personal nouns. Lexical gender may or may not be marked morphological-
ly. In English, most human nouns are not formally marked for lexical gender,
with exceptions such as widow–widower or steward–stewardess, which show
overt gender marking by suffixation. Only in principle is such markedness
independent of grammatical gender. Languages with grammatical gender
generally possess a much larger number of devices of overt gender marking.
Thus, in the highly inflected Slavic languages, overt lexical gender marking (as
a result of the correspondence with grammatical gender) is much more visible
than in most Germanic languages, simply because satellite elements have more
gender-variable forms.
3.3 Referential gender

“Referential gender” relates linguistic expressions to the non-linguistic reality;
more specifically, referential gender identifies a referent as “female”, “male” or
“gender-indefinite”. For example, a personal noun like Germ. Mädchen ‘girl’ is
grammatically neuter, has a lexical-semantic specification as [female], and is
generally used to refer to females. However, an idiomatic expression like
Mädchen für alles lit. ‘girl for everything’; ‘maid of all work’, may be used to
refer to males also. In this example, while the metaphor seems to neutralize the
lexical specificity of Mädchen, a gendered message is nevertheless transmitted:
the expression has explicitly derogatory connotations.
In gender languages, a complex relationship between grammatical gender

and referential gender obtains for the majority of personal nouns, with typical
gender-related asymmetries in pronominalization and coordination (cf.
Sections 4.3 and 4.4 below). For example, when reference is made to a particu-
lar known individual, the choice of anaphoric pronouns may be referentially
motivated and may thus override the noun’s grammatical gender, as in Germ.
Tennisstar (m) … sie (f) (cf. Oelkers 1996).
3.4 “False generics”: Generic masculines and male generics

All the gender languages of the project illustrate the traditional (and often
prescriptive) practice which requires the use of so-called “generic masculines”
to refer to males as well as females.13 With reference to languages with gram-
matical gender we will talk about “generic masculines” (where “masculine”
denotes grammatical gender), while for languages without grammatical gender,
such as English or Japanese, the term “male generics” (with “male” denoting a
lexical-semantic property) is more appropriate. This terminological distinction
reflects on the different typological affiliations of the respective languages as
explained in Section 2.
Grammatically feminine personal nouns tend to be female-specific (with
only few exceptions), while grammatically masculine nouns have a wider lexical
and referential potential. For example, masculine nouns such as Russ. vrač (m)
‘physician’, Fr. ministre (m) ‘minister’, or Arab. muAami (m) ‘lawyer’ may be
used to refer to males, groups of people whose gender is unknown or unimpor-
tant in the context, or even female referents, illustrating the function of the so-
called “generic masculine” usage. The reverse, i.e. the use of feminine nouns
with gender-indefinite reference, is the rare exception. For example, in Seneca,
an Iroquoian language, the feminine has been attested for indefinite reference
to people in general (cf. Chafe 1967). In Oneida, also an Iroquoian language,
gender-indefinite reference may be achieved by feminine pronouns. But then,
speakers may make other choices (including the masculine gender) which are
determined by highly complex semantic and pragmatic constraints (cf. Abbott
1984: 126). In a number of Australian Aboriginal languages, the feminine is
used as the unmarked gender – in restricted contexts –, while other languages
from the same family exhibit the opposite configuration (Alpher 1987: 175).
Clearly, further research is necessary which must bring together the anthropo-
logical and linguistic evidence. Of primary importance will be the question in
which way a relationship can be identified between the existence of femi-
nine/female generics and underlying matriarchal structures.
In languages without grammatical gender, but with some gender-variable

pronouns, male generic usage is the traditional androcentric practice in cases of
gender-indefinite reference. E.g., in English, where gendered third person
singular pronominal distinctions remain of an original grammatical gender
system, “generic he” – including him(self) – is the prescriptive choice in such
cases as an American drinks his coffee black. Since the use of male-biased
pronouns may create referential ambiguities and misunderstandings, alternative
formulations have been suggested to replace male generic expressions, e.g.,
Americans drink their coffee black (cf. Section 6). In languages without pronomi-
nal gender distinctions, male generic usage is found with the nouns themselves.
In Finnish, for example, occupational terms ending in -mies ‘man’ are used for
men as well as women (e.g., lakimies lit. ‘law-man’; ‘lawyer’) and are officially
claimed to be gender-neutral. Empirical findings reported by Engelberg (this
vol.), however, show that this claim is more than doubtful.
The prescription of “generic masculines” or “male generics” has long been
the center of debates about linguistic sexism in English and other languages.
The asymmetries involved here, i.e. the choice of masculine/male expressions as
the normal or “unmarked” case with the resulting invisibility of feminine/
female expressions are reflections of an underlying gender belief system, which
in turn creates expectations about appropriate female and male behavior. Such
expectations will prevent a genuinely generic interpretation of gender-indefinite
personal nouns, and can also be related to the fact that masculine/male pro-
nouns occur three times as frequently as the corresponding feminine/female
pronouns in some languages, e.g., in English and Russian.14 There is empirical
evidence for English, but also for Turkish, Finnish, and German, that most
human nouns are in fact not neutral, which supports the assumption that
gender-related socio-cultural parameters are a powerful force in shaping the
semantics of personal reference.15
3.5 Social gender

“Social gender” is a category that refers “to the socially imposed dichotomy of
masculine and feminine roles and character traits” (Kramarae & Treichler
1985: 173). Personal nouns are specified for social gender if the behavior of
associated words can neither be explained by grammatical nor by lexical gender.
An illustration of social gender in English is the fact that many higher-status
occupational terms such as lawyer, surgeon, or scientist will frequently be
pronominalized by the male-specific pronoun he in contexts where referential
gender is either not known or irrelevant. On the other hand, low-status

occupational titles such as secretary, nurse, or schoolteacher will often be
followed by anaphoric she. But even for general human nouns such as pedestri-
an, consumer or patient, traditional practice prescribes the choice of he in
neutral contexts.
Social gender has to do with stereotypical assumptions about what are
appropriate social roles for women and men, including expectations about who
will be a typical member of the class of, say, surgeon or nurse. Deviations from
such assumptions will often require overt formal markings, as in Engl. female
surgeon or male nurse. However, since the majority of general personal nouns
can be assumed to have a male bias, it seems plausible to suggest that – irrespec-
tive of whether the language does or does not have grammatical gender –
underlying is the principle “male as norm”.
Social gender is a particularly salient category in a language like Turkish
which lacks even gender-variable pronouns. Frequently, gender-related associa-
tions remain hidden on a deeper semantic level. E.g., the Turkish occupational
term kuyumcu ‘goldseller’ is lexically gender-indefinite, but is invariably
associated with male referents, although theoretically, a female goldseller could
also be referred to as kuyumcu. The word can be said to have a covert male bias
which derives from sociocultural assumptions and expectations about the
relationships between women and men (cf. Braun, vol. I, Section 3.1).
4. Gender-related structures
4.1 Word-formation
Word-formation is a particularly sensitive area in which gender may be commu-
nicated. In languages with or without grammatical gender, processes of deriva-
tion and compounding have an important function in the formation of gendered
personal nouns, particularly in the use of existing and the creation of new femi-
nine/female terms, e.g., in the area of occupational terms, cf. (1) and (2):
(1) Derivation
Masculine/male Feminine/female
Norw. forfatter forfatter-inne ‘author’
Arab. katib katib-a ‘secretary’
Rom. pictor pictor-iţă ‘painter’
Engl. steward steward-ess
(2) Compounding
Masculine/male Feminine/female
Germ. Geschäfts-mann Geschäfts-frau ‘business man/woman’
Norw. politi-mann politi-kvinne ‘police officer’
lit. ‘police man’ lit. ‘police woman’
EMC seli-man seli-uman ‘trader’
lit. ‘sell-man’ lit. ‘sell-woman’
Indon. dokter dokter perempuan ‘doctor’
lit. ‘doctor’ lit. ‘doctor woman’
Typically, female gender-specification occurs with reference to a particular

individual (Congresswoman Maxine Waters) or in contexts of contrastive
emphasis (male and female delegates). Female linguistic visibility is often a
marked and loaded concept, and we find considerable variation concerning the
status and productivity of feminine/female word-formation processes across
languages. Thus, German has a well-established and extremely productive
process for the formation of personal feminines ending in -in: Punkerin ‘female
punk’, Bundeskanzlerin ‘female chancellor’, Bischöfin ‘female bishop’, etc. By
contrast, Welsh, also a gender language, has no such instrument for morpho-
logical gender-specification. Very few derived feminines exist, i.e. most occupa-
tional and other personal nouns in Welsh are grammatically masculine and
have no feminine counterparts.
In English, the few derivational patterns that exist for the formation of
female-specific terms have low productivity, and more often than not produce
semantically asymmetric pairs in which the female represents the lesser catego-
ry, illustrating what Schulz (1975) has called “semantic derogation”. Notorious
examples are Engl. governor/governess, major/majorette. Of course, such asym-
metric pairs also occur in languages with grammatical gender, cf. (3):
(3) Fr. couturier (m) ‘fashion designer’
couturière (f) ‘seamstress, female tailor’
Germ. Sekretär (m) ‘secretary of an administration, trade union
or the like’
Sekretärin (f) ‘secretary in an office’
Feminine/female terms are not consistently derived nor used in case of female
reference; their use may be stylistically marked and in many languages carries
negative connotations, which makes them unacceptable in neutral contexts.
Thus, in Russian or Polish, where masculinity is highly valued, feminine/female
counterparts of terms denoting prestigious occupations are avoided. By contrast,

masculine/male terms are either neutral or carry positive connotations.
4.2 Agreement
In agreement, concern is with overt representations of gender. On a formal
level, agreement establishes a syntactic relationship between a noun’s satellite
element, e.g., an article, adjective, pronominal or verbal form, and the noun’s
gender class. Satellite elements must be gender-variable, i.e. they must allow for
a choice between at least two values (e.g., feminine and masculine, as in French
and Italian, or feminine, masculine and neuter, as in Russian and German). In
some languages, e.g., in Russian, discourse categories such as the gender of
speaker, addressee or person talked about may all be marked morphologically
on some verbal forms (cf. Doleschal & Schmid, vol. I, Section 2.2):
(4) Prišl-a moj-a byvš-aja studentka,
came-fem my-fem former-fem student.fem
kotor-aja očen’ umn-aja. On-a mogl-a by pomoč’.
who-fem very intelligent-fem she-fem might-fem cond help
‘A former student of mine, who is very intelligent, has come. She
might help.’
In traditional grammars, agreement is described as a primarily formal and

predictable phenomenon, one of whose major functions is reference-tracking.
Contrary to this view, we believe that agreement may add semantic and social
information to the discourse, thus taking on symbolic functions. This claim is
based on the observation that agreement tends to affect masculine and feminine
nouns in different ways, mainly due to the principle “male as norm”: Agree-
ment will favor the masculine in coordination (cf. Section 4.4), and, generally,
masculine agreement predominates; feminine agreement is female-specific and,
in many contexts, non-obligatory and irregular, depending on extralinguistic
factors such as tradition, prescription or speaker attitude.
4.3 Pronominalization
Gendered pronouns are overt representations of gender both in languages with
and without grammatical gender. Anaphoric gendered pronouns reveal the
semantic specification of nouns with lexical gender, they may express referential
gender in contradiction to grammatical gender, they may function as a means
to either specify or abstract from (intended) referential gender, and they may
emphasize traditional or reformed practices, as when a speaker chooses between
a “false generic” (e.g., Engl. he) or a more gender-neutral alternative (such as
Engl. “singular they”). Generally, pronominalization is a powerful strategy of
communicating gender.
The interpretation of pronominalization as one type of agreement remains
controversial. English exemplifies a type of relation between noun and pronoun
which is not syntactically motivated. Only reflexes of the original grammatical
gender system remain in third person singular pronouns (he–she–it), and the
choice of anaphoric pronouns is controlled by lexical-semantic properties of the
antecedent, by referential gender (including intended reference), or social
gender. Corbett (1991: 169) concludes that pronouns “may be the means by
which particular languages divide nouns into different agreement classes”.
However, this classification is semantically based, and English is, of course, not
a “gender language” as defined in Section 2.2.
4.4 Coordination
When a noun phrase conjoins a masculine and a feminine noun, the choice of
a related target form may create a conflict between two competing genders. An
example from Romanian (cf. Maurice, vol. I, Section 2.3) illustrates the strategy of
what Corbett (1991: 279) calls “syntactic gender resolution”, where agreement
occurs with one conjunct only, namely the masculine, albeit in the plural:16
(5) un vizitator şi o turistă mult interesaţi
a visitor.masc and a tourist.fem very interested.masc.pl
‘a very interested (male) visitor and a very interested (female) tourist’
Corbett claims that the choice of masculine agreement forms in such cases is
“evidently of the syntactic type” (Corbett 1991: ibid.), since what determines
agreement is independent of the meaning of the nouns involved. In our view,
however, the example illustrates the prescriptive practice that if at least one
conjunct is headed by a masculine noun, masculine agreement forms are used.
Another illustration of this practice involving inanimate nouns is the Hebrew
example (6), cf. Tobin (vol. I, Section 2.3):
(6) Ha-sefer ve-ha-maxberet nimtsaim kan.
the-book.masc.sg and-the-notebook.fem.sg are.found.masc.pl here
‘The book and the notebook are here.’
There are a number of exceptions to this regularity. For example, in some

languages with three grammatical genders, the neuter gender may be employed
to resolve the gender conflict in coordination, as in this example from Icelandic
(cf. Grönberg, this vol., Section 2.3):
(7) Óli og Elsa eru ung.
Óli.masc and Elsa.fem are young.neut.pl
‘Óli and Elsa are young.’
In some cases the choice of the masculine target gender may be motivated by
the vicinity of the nearest controller noun when this is also masculine (cf.
Corbett 1991: 265). However, “Gender across languages” provides numerous
counter-examples. For example, in Arabic, if word order in a conjoined noun
phrase is reversed to masculine first and feminine second, the choice of the
feminine, as a response to the nearest controller gender, is ungrammatical; the
masculine must still be chosen (cf. Hachimi, vol. I, Section 4.3):
(8) Lab u bnat-u yan-in.
father.masc.sg and daughter.fem.pl-his tired.masc-pl
‘The father and his daughters are tired.’
Underlying such syntactic conventions may be a gender hierarchy which defines

the masculine as the “most worthy gender” (Baron 1986: 97).17 As a result,
masculine nouns are highly visible in gender languages and carry considerably
more weight and emphasis than feminine nouns.
5. Gender-related messages
The communication of gender-related messages may be performed by many

other devices in addition to the ones discussed so far. Of primary importance in
the context of “Gender across languages” are address forms, idiomatic and
metaphorical expressions, proverbs, and, of course, female/male discourse.
5.1 Address terms

Languages differ considerably in the type of obligatory and optional informa-
tion they encode in their address systems. English can be characterized as a
language with only moderate distinctions, lacking even the tu/vous-distinction
that is characteristic, e.g., for German, French or Russian, while languages such
as Vietnamese, Japanese or Javanese have extremely complex address systems.18
For example, on the basis of the underlying, all-pervasive concept of

hormat ‘respect’, Indonesian as spoken in Java has lexicalized numerous socio-
cultural and interactional dimensions such as age, gender, social status,
participant relationship, and formality of the situation, which will determine
a speaker’s selection of an item from one of several speech styles and terms (cf.
Kuntjara, vol. I, Section 3). Gender will be performed in asymmetric and non-
reciprocal practices. Thus, the traditional Javanese husband will address his
wife by her first name or by the kinship term dik ‘younger sister’, but will
receive the term mas ‘older brother’, irrespective of his age. Lexical choices
generally are less constrained for males, while women are expected to use a
higher, more deferential style.
Changes in address practices may be indicative of underlying changes in the
social relationships between women and men. In language planning such
changes will be supported as contributing to more symmetry in address
systems. An example is the legislation establishing Germ. Frau as the only
acceptable official term of address for adult women to abolish the traditional
distinction between Frau ‘Mrs’ and Fräulein ‘Miss’ (cf. Bußmann & Hellinger,
vol. III).19 Similarly, in English the address term Ms was introduced to abolish
the distinction between Mrs and Miss. However, such a term may also be
appropriated by mainstream usage to transmit (originally) unintended messag-
es, as in the case of Australian English Ms ‘divorced’ or ‘feminist’ (cf. Pauwels,
vol. I, Section 2.1).
5.2 Idiomatic expressions and proverbs

Another area of the implicit discursive negotiation of gender, irrespective of
whether the language does or does not have grammatical gender, are frozen
expressions such as idioms, metaphors, and proverbs.20 Descriptions of or terms
for women – when these are part of such expressions – tend to have negative,
and frequently sexual and moral implications which are not found for corre-
sponding male terms (where these exist).
For example, Moroccan Arabic provides a number of honorific terms,
phrases, and proverbs which are indicative of the glorification of the mother-
concept in Moroccan culture, as in ‘the mother is the light of the house’ or
‘paradise lies under mothers’ feet’. At the same time, mothers of daughters are
evaluated negatively, reflecting on the unequal status of girls and boys (cf.
Hachimi, vol. I, Section 7). Representative of the genre of proverbs is the
following Turkish example (cf. Braun, vol. I, Section 6):
(9) Oglan doguran övünsün, kiz doguran dövünsün.

‘Let the one who bears a son be proud, let the one who bears a daughter
beat herself.’
This is the message of numerous idiomatic expressions and proverbs from

many languages of “Gender across languages”: Arabic, Chinese, Danish,
Finnish, Italian, Norwegian, Russian, and Turkish.
In Russian, the woman-as-mother concept is practically the only positive
female image in proverbs (cf. Doleschal & Schmid, vol. I, Section 6.1). The
extreme opposite is obscene language with expressions of “mother-fucking”, a
misogynist practice which has also been attested for many languages, with
Russian, Chinese, Turkish, and Danish representing examples in “Gender
across languages”. Such frozen expressions embody fundamental collective
beliefs and stereotypes which are available for continued practices of communi-
cating gender.
5.3 Female and male discourse

A major concern of studies on language and gender in the 1990s has been the
search for an empirical foundation on which statements could be made on
discourse practices in diverse types of interaction (cf. Wodak & Benke 1997).
On a theoretical level the inadequacy of binary categories (women vs. men,
female vs. male) has been criticized. These categories show internal diversifica-
tion and must be described to a considerable extent as social constructs. Also
statements about female and male verbal behavior have been criticized for
making inappropriate generalizations. Explanatory theories (cf. the deficit,
dominance, difference, and diversity models) developed with reference to
English cannot be applied to other languages without taking into account
dimensions of sociocultural difference (cf. also Pauwels 1998, Bergvall 1999).
Investigations of gender and discourse have primarily focussed on the
identification of differences between female and male speech.21 For a number
of languages, among them English, Chinese and Japanese, some differences
were indeed found, but quantitative evidence remains controversial. For
example, higher frequencies of “uncertainty phenomena” were found in some
types of discourse (typically in experimental or more formal situations), but not
in others. More importantly, the occurrence of tag-questions (e.g. in English)
or sentence-final particles (e.g. in Chinese) may have various communicative
functions in actual discourse, so that an explanation in terms of uncertainty or
tentativeness is only one among several possibilities (cf. Holmes 1995). This is
also true for categories of turn-taking, where a higher frequency of interrup-
tions and overlaps as performed by male speakers is widely interpreted as
indicative of conversational dominance (cf. West & Zimmerman 1983).
However, Bergvall (1999) has repeatedly warned against immediately approach-
ing discourse in terms of gender differences, suggesting that rather than
categorizing people and their verbal behavior into seemingly dichotomous and
opposed groups, it would be more appropriate to interpret the data in terms of
a linguistic and behavioral continuum.
In “Gender across languages”, discourse analysis features more prominently
for those languages where – in the absence of substantial structural representa-
tions of gender – discourse emerges as a central field in which gender is
negotiated, e.g., in Chinese, Japanese, English, and Belizean Creole.
6. Language change and language reform
In all the languages represented in “Gender across languages”, tendencies of

variation and change in the area of personal reference can be observed. In some
languages (e.g., English, German, French, Dutch and Spanish) such tendencies
have been supported by language planning measures, including the publication
of recommendations and guidelines, while for other languages an awareness of
gendered asymmetries is only beginning to develop in both academia and the
media (e.g., in Czech or Polish). To a large extent, the emergence of public
discourse on language and gender depends on the socio-political background,
in particular the state of the women’s movement in the respective country.
Language as a tool of social practice may serve referential functions (e.g.,
the exchange of information); it has social-psychological functions in that it
reflects social hierarchies and mechanisms of identification, and it contributes
to the construction and communication of gender. More specifically, language
is assumed to codify an androcentric worldview. Recommendations and
guidelines for non-discriminatory language identify areas of conventional
language use as sexist and offer alternatives aiming at a gender-fair (and
symmetric) representation of women and men. As an instrument of language
planning they reinforce tendencies of linguistic change by means of explicit
directions (cf. Frank 1989: 197; Pauwels 1998, 1999; Hellinger 1995).
Gender-related language reform is a reaction to changes in the relationships
between women and men, which have caused overt conflicts on the level of
language comprehension and production. Reformed usage symbolizes the

dissonance between traditional prescriptions such as the use of masculine/male
generics and innovative alternatives. In most cases it explicitly articulates its
political foundation by emphasizing that equal treatment of women and men
must also be realized on the level of communication.
Guidelines are based on the assumption that a change in behavior, i.e.,
using more instances of non-sexist language, will be attended by a change in
attitude so that positive attitudes towards non-sexist alternatives will develop
(cf. Smith 1973: 97). Conversely, positive attitudes will motivate speakers to use
more non-sexist language. This is not necessarily what happens in actual cases
of language reform. Reformed usage has sometimes been appropriated by
speakers who will use alternatives in ways that were not intended, thereby
redefining and depoliticizing feminist meanings (cf. Ehrlich & King 1994).
7. Conclusion
The central function of linguistic gender in the domain of human reference is

the communication of gendered messages of various types. The linguistic
representation of gender is one of the dimensions on which languages can be
compared, irrespective of individual structural properties and sociolinguistic
diversities. However, even apparently straightforward categories such as
grammatical or referential gender cannot be fully described in terms that
abstract from the cultural and sociopolitical specifics of individual languages.
And once the study of gender is taken beyond the level of formal manifestation
to include discourse practices, the concept of gender becomes increasingly
complex and multi-dimensional.
The general tendencies we have identified all center around one fundamen-
tal principle: masculine/male expressions (and practices) are the default choice
for human reference in almost any context. The assumption may be plausible
that gender languages offer the larger potential for the avoidance of male-biased
language – simply because female visibility is more easily achieved on the level
of expression. At the same time, advocating an increase in female visibility may
create problematic and potentially adverse effects in languages like Russian or
Hebrew, where masculine/male terms for female reference are evaluated
positively even by women. In addition, consistent splitting, i.e. the explicit use of
both feminine and masculine expressions when reference is made to both women
and men, is considered to be stylistically cumbersome by many speakers, esp. in
languages with case. Thus, a comparative view would have to investigate the ways
in which structural prerequisites interact with sociolinguistic tendencies of change.
By contrast, “genderless” languages seem to provide more possibilities for
egalitarian and gender-neutral expressions, by avoiding the dominant visibility
of masculine terms, and stereotypical associations of feminine terms with
secondary or exceptional status. However, in genderless languages it may be
even more difficult to challenge the covert male bias and the exclusion of female
imagery in many personal nouns.
In the study of language and gender, there is an urgent need for compara-
tive analyses based on adequate descriptions of a large number of languages of
diverse structural and sociocultural backgrounds. This includes an awareness
of the fact that white middle class North American English cannot be regarded
as representative for other languages also. “Gender across languages” contrib-
utes towards the goal of a more global view of gender by presenting a wealth of
data and language-specific analyses that will allow for cross-linguistic state-
ments on manifestations of gender. In addition, the material presented in
“Gender across languages” can be expected to enrich the debate of a number
of interdisciplinary issues:
From a sociolinguistic perspective, the tremendous variation found in the
exchange of gendered messages must be placed more explicitly in a wider
framework of communities of practice (CofP), considering the interaction
between “gender” and age, ethnic membership, social status and religion.22
From a text-linguistic perspective, comparative investigations of gender-
related structures will identify the stylistic and rhetorical potentials of grammat-
ical gender in a given language, in particular for the construction of cohesion
and textuality by a less constrained word order and for disambiguation (refer-
ence tracking).
From a historical perspective, the analysis of ongoing structural changes
may shed light on the question of why manifestations of gender in historically
or typologically related languages have developed in very different directions, as
in the case of Germanic languages which may have two or three categories of
grammatical gender – or none at all.
From a psycholinguistic perspective, further empirical evidence is needed
from more languages that might contribute towards an understanding of how
gendered messages are interpreted, and more generally, in which ways the
perception and construction of the universe is influenced by linguistic, social
and cultural parameters.
<DEST "hel-n*">
Notes
* In accordance with the publisher’s policy of publishing “Gender across languages” as three
separate and independent volumes, this chapter is a reprint from volume I. Of course,
references to chapters in the other two volumes have been adjusted.
1. Cf. Corbett (1991). Lehmann (1993) provides an informative overview of types of
congruence/agreement. Rich data from various languages can be found in Barlow &
Ferguson (1988).
2. Cf. Zubin & Köpcke (1984, 1986).
3. Cf. also Lakoff (1987: ch. 6), Corbett (1991: 15–18). For further examples see Grimm
(1831: 349f.), Royen (1929: 341–347), Strunk (1994: 151f).
4. On the origin of gender cf. Claudi (1985), Fodor (1959), Ibrahim (1973), Royen (1929),
Leiss (1994); on the decay and loss of gender (systems) cf. Corbett (1995), Claudi (1985).
5. This ignores the very rudimentary numeral classification found in Turkish.
6. Cf. Unterbeck (2000) for an overview of different types of noun classification. Material
from a larger number of languages can be found in Craig (1986, 1994). Royen (1929) is still
an impressive study of gender and nominal classification.
7. Thus, for Vietnamese over 200 such classifiers have been identified, cf. Pham (this vol.,
Section 2); on classifier languages cf. also Craig (1994).
8. For example, Corbett (1991: ch. 3.1) discusses morphological gender assignment jointly for
Russian, Swahili and other Bantu languages; cf. also Hurskainen (2000).
9. This is the approach taken by Dixon (1982: 160); cf. also Braun (2000: 32).
10. Swahili (cf. Beck, vol. III) is one of perhaps 600 African languages with noun classes (cf.
Heine 1982:190); on noun classes in African languages cf. Hurskainen (2000). Large numbers
of noun class languages are also found among Dravidian and New Guinean languages.
11. In contrast to English, Persian even lost pronominal gender distinctions. The loss of
grammatical gender in English is described in Jones (1988), and more recently, Kastovsky
(2000); for a diachronic perspective on gender in the Scandinavian languages cf. Braunmüller
(2000), in French cf. Härmä (2000), and in the Iranian languages Corbett (1991: 315–318).
12. Practices of gender-crossing in Native American communities, e.g., the Navajo, are
described in Whitehead (1991). So-called “abnormal” developments are discussed in Wodak
& Benke (1997: ch. 1.2).
13. The term “false generics” was used by Kramarae & Treichler (1985: 150, 175) to refer to
“generic masculines”. Romaine (vol. I, Section 3.2) uses the term “androcentric generics”.
14. There are statistical data for English (Graham 1975) and Russian (Francis & Kučera 1967).
15. Empirical evidence for English can be found in MacKay & Fulkerson (1979), for Turkish
in Braun (2000), for Finnish in Engelberg (this vol., Section 5), for German in Scheele &
Gauler (1993) and Irmen & Köhncke (1996). For cross-linguistic evidence cf. Batliner (1984).
16. Coordination is no problem in German which has no corresponding gender-variable
satellite forms in the plural (cf. Bußmann & Hellinger, vol. III).
17. Cf. also Curzan (2000); for German, cf. Bußmann (1995).
18. On address systems, cf. Braun (1988); on the T/V distinction Brown & Gilman (1960).
For Vietnamese, cf. Pham, this volume.
19. On French legislation, cf. Burr, vol. III.
20. For German, cf. Daniels (1985), for Moroccan cf. Webster (1982), for Chinese cf. Zhang
(this vol.). For a comparison of Finnish and German proverbs cf. Majapuro (1997).
21. For recent overviews of gendered discourse, cf. Talbot (1998) and Romaine (1999: chs. 6, 7).
22. On the concept of CofP, cf. the special issue of Language in Society 28/2 (1999).
References
Abbott, Clifford. 1984. “Two feminine genders in Oneida.” Anthropological Linguistics 26:
125–137.
Alpher, Barry. 1987. “Feminine as the unmarked grammatical gender: Buffalo girls are no
fools.” Australian Journal of Linguistics 7: 169–187.
Barlow, Michael & Charles A. Ferguson, eds. 1988. Agreement in natural language. Approach-
es, theories, descriptions. Stanford, CA: Stanford University Press.
Baron, Dennis. 1986. Grammar and gender. New Haven, CT: Yale University Press.
Batliner, Anton. 1984. “The comprehension of grammatical and natural gender: A cross-
linguistic experiment.” Linguistics 22: 831–856.
Bergvall, Victoria L. 1999. “An agenda for language and gender research for the start of the
new millennium.” Linguistik Online. http://viadrina.euv-frankfurt-o.de/~wjournal/
1_99/bergvall.htm
Bergvall, Victoria L. & Janet M. Bing & Alice F. Freed, eds. 1996. Rethinking language and
gender research: Theory and practice. London: Longman.
Bing, Janet M. & Victoria L. Bergvall. 1996. “The question of questions: Beyond binary
thinking.” In Rethinking language and gender research: Theory and practice, eds. Victoria
L. Bergvall & Janet M. Bing & Alice F. Freed. London: Longman, 1–30.
Bittner, Dagmar. 2000. “Gender classification and the inflectional system of German.” In
Gender in grammar and cognition, eds. Barbara Unterbeck et al. Berlin: de Gruyter, 1–23.
Braun, Friederike 1988. Terms of address. Problems of patterns and usage in various languages
and cultures. Berlin: Mouton de Gruyter.
Braun, Friederike. 2000. Geschlecht im Türkischen. Untersuchungen zum sprachlichen Umgang
mit einer sozialen Kategorie. Wiesbaden: Harrassowitz.
Braunmüller, Kurt. 2000. “Gender in North Germanic: A diasystematic and functional ap-
proach.” In Gender in grammar and cognition, eds. Barbara Unterbeck et al. Berlin: de
Gruyter, 25–54.
Brown, Roger & Albert Gilman. 1960. “The pronouns of power and solidarity.” In Style in
language, ed. Thomas A. Sebeok. Cambridge, MA: MIT Press, 253–276.
Bußmann, Hadumod. 1995. “Das Genus, die Grammatik und – der Mensch: Geschlechter-
differenz in der Sprachwissenschaft.” In Genus: Zur Geschlechterdifferenz in den Kultur-
wissenschaften, eds. Hadumod Bußmann & Renate Hof. Stuttgart: Kröner, 114–160.
Chafe, Wallace L. 1967. Seneca morphology and dictionary. Washington: Smithsonian Press.
Claudi, Ulrike. 1985. Zur Entstehung von Genussystemen: Überlegungen zu einigen theore-
tischen Aspekten, verbunden mit einer Fallstudie des Zande. Hamburg: Buske.
Corbett, Greville G. 1991. Gender. Cambridge: Cambridge University Press.
Corbett, Greville G. 1995. “Agreement.” In Syntax. Ein internationales Handbuch zeit-
genössischer Forschung, Vol. 2, eds. Joachim Jacobs & Arnim von Stechow & Wolfgang
Sternefeld & Theo Vennemann. Berlin: de Gruyter, 1235–1244.
Corbett, Greville G. & Norman M. Fraser. 2000. “Default genders.” In Gender in grammar
and cognition, eds. Barbara Unterbeck et al. Berlin: de Gruyter, 55–97.
Craig, Colette G. 1986. Noun classes and categorization. Amsterdam: Benjamins.
Craig, Colette G. 1994. “Classifier languages.” In The encyclopedia of language and linguistics,
Vol. 2, ed. Ronald E. Asher. Oxford: Pergamon, 565–569.
Curzan, Anne. 2000. “Gender categories in early English grammars: Their message to the
modern grammarian.” In Gender in grammar and cognition, eds. Barbara Unterbeck et
al. Berlin: de Gruyter, 561–576.
Daniels, Karlheinz. 1985. “Geschlechtsspezifische Stereotypen im Sprichwort.” Sprache und
Literatur 56: 18–35.
Dixon, Robert M. W. 1972. The Dyirbal language of North Queensland. Cambridge: Cam-
bridge University Press.
Dixon, Robert M.W. 1982. Where have all the adjectives gone? And other essays in semantics
and syntax. Berlin: Mouton.
Ehrlich, Susan & Ruth King. 1994. “Feminist meanings and the (de)politicization of the
lexicon.” Language in Society 23: 59–76.
Fodor, Istvan. 1959. “The origin of grammatical gender.” Lingua 8: 1–41, 186–214.
Francis, Nelson & Henry Kučera. 1967. Computational analysis of present-day American
English. Providence, RI: Brown University Press.
Frank, Francine W. 1989. “Language planning, language reform, and language change: A
review of guidelines for nonsexist usage.” In Language, gender, and professional writing:
Theoretical approaches and guidelines for nonsexist usage, eds. Francine W. Frank & Paula
A. Treichler. New York: MLA, 105–137.
Graham, Alma. 1975. “The making of a nonsexist dictionary.” In Language and sex: Differ-
ence and dominance, eds. Barrie Thorne & Nancy Henley. Rowley, MA: Newbury
House, 57–63.
Greenberg, Joseph H. 1972. “Numeral classifiers and substantival number: Problems in the
genesis of a linguistic type.” Working Papers on Language Universals 9: 1–39.
Greenberg, Joseph H. 1978. “How does a language acquire gender markers?” In Universals of
human language, Vol. 3, ed. Joseph H. Greenberg. Stanford, CA: Stanford University
Press, 47–82.
Grimm, Jacob. 1831. Deutsche Grammatik III. 3. Theil, 3. Buch: Von der Wortbildung.
Göttingen: Dietrich.
Härma, Juhani. 2000. “Gender in French: A diachronic perspective.” In Gender in grammar
Hall, Kira & Mary Bucholtz, eds. 1995. Gender articulated: Language and the socially con-
structed self. New York: Routledge.
Heine, Bernd. 1982. “African noun class systems.” In Apprehension: Das sprachliche Erfassen
von Gegenständen. Teil I: Bereich und Ordnung der Phänomene, eds. Hansjakob Seiler &
Christian Lehmann. Tübingen: Narr, 189–216.
Hellinger, Marlis. 1995. “Language and gender.” In The German language in the real world,
ed. Patrick Stevenson. Oxford: Clarendon, 280–314.
Hockett, Charles F. 1958. A course in modern linguistics. New York: Macmillan.
Holmes, Janet. 1995. Women, men and politeness. London: Longman.
Hurskainen, Arvi. 2000. “Noun classification in African languages.” In Gender in grammar
Ibrahim, H. Muhammad. 1973. Grammatical gender. Its origin and development. The Hague:
Mouton.
Irmen, Lisa & Astrid Köhncke. 1996. “Zur Psychologie des “generischen” Maskulinums.”
Sprache & Kognition 15: 152–166.
Jones, Charles. 1988. Grammatical gender in English: 950–1250. London: Croom Helm.
Kastovsky, Dieter. 2000. “Inflectional classes, morphological restructuring, and the dissolu-
tion of Old English grammatical gender.” In Gender in grammar and cognition, eds.
Barbara Unterbeck et al. Berlin: de Gruyter, 709–727.
Kramarae, Cheris & Paula A. Treichler. 1985. A feminist dictionary. Boston: Pandora.
Lakoff, George. 1987. Women, fire, and dangerous things: What categories reveal about the
mind. Chicago: University of Chicago Press.
Lehmann, Christian. 1993. “Kongruenz.” In Syntax. Ein internationales Handbuch zeit-
genössischer Forschung, Vol.1, eds. Joachim Jacobs & Arnim von Stechow & Wolfgang
Sternefeld & Theo Vennemann. Berlin: de Gruyter: 722–729.
Leiss, Elisabeth. 1994. “Genus und Sexus. Kritische Anmerkungen zur Sexualisierung von
Grammatik.” Linguistische Berichte 152: 281–300.
MacKay, Donald & David Fulkerson. 1979. “On the comprehension and production of
pronouns.” Journal of Verbal Learning and Verbal Behavior 18: 661–673.
Majapuro, Anne. 1997. “Weinende Braut, lachende Frau. Geschlechtsspezifische Merkmale
des Lachens und Weinens in deutschen und finnischen Sprichwörtern.” In Laughter down
the centuries, eds. Siegfried Jäckel & Asko Timonen. Turku: University of Turku, 233–248.
Oelkers, Susanne. 1996. “ ‘Der Sprintstar und ihre Freundinnen’. Ein empirischer Beitrag zur
Diskussion um das generische Maskulinum.” Muttersprache 106: 1–15.
Pauwels, Anne. 1998. Women changing language. London: Longman.
Pauwels, Anne. 1999. “Feminist language planning: Has it been worthwhile?” Linguistik
Online. http://www.viadrina.euv-frankfurt-o.de/~wjournal/1_99/pauwels.htm
Romaine, Suzanne. 1999. Communicating gender. Mahwah, NJ: Lawrence Erlbaum.
Royen, Gerlach. 1929. Die nominalen Klassifikations-Systeme in den Sprachen der Erde.
Historisch-kritische Studie, mit besonderer Berücksichtigung des Indo-Germanischen.
Mödling: Anthropos.
Scheele, Brigitte & Eva Gauler. 1993. “Wählen Wissenschaftler ihre Probleme anders aus als
WissenschaftlerInnen? Das Genus-Sexus-Problem als paradigmatischer Fall der
linguistischen Relativitätsthese.” Sprache & Kognition 12: 59–72.
Schulz, Muriel R. 1975. “The semantic derogation of woman.” In Language and sex:
Difference and dominance, eds. Barrie Thorne & Nancy Henley. Rowley, MA: Newbury
House, 64–75.
</TARGET "hel">
Smith, David M. 1973. “Language, speech and ideology: A conceptual framework.” In

Language attitudes: Current trends and prospects, eds. Roger W. Shuy & Ralph W. Fasold.
Washington, DC: Georgetown University Press, 97–112.
Strunk, Klaus. 1994. “Grammatisches und natürliches Geschlecht in sprachwissenschaftlicher
Sicht.” In Frau und Mann. Geschlechterdifferenzierung in Natur und Menschenwelt, ed.
Venanz Schubert. St. Ottilien: EOS Verlag, 141–164.
Talbot, Mary M. 1998. Language and gender: An introduction. Oxford: Blackwell.
Unterbeck, Barbara. 2000. “Gender: New light on an old category. An introduction.” In Gender
in grammar and cognition, eds. Barbara Unterbeck et al. Berlin: de Gruyter, xv–xlvi.
Webster, Sheila K. 1982. “Women, sex, and marriage in Moroccan proverbs.” International
Journal of Middle East Studies 14: 173–184.
West, Candace & Don H. Zimmerman. 1983. “Small insults: A study of interruptions in
cross-sex conversations between unacquainted persons.” In Language, gender and
society, eds. Barrie Thorne & Cheris Kramarae & Nancy Henley. Rowley, MAU:
Newbury House, 113–118.
Whitehead, Harriet. 1991 [11981]. “The bow and the burden strap: A new look at institu-
tionalized homosexuality in Native North America.” In Sexual meanings. The cultural
construction of gender and sexuality, eds. Sherry B. Ortner & Harriet Whitehead.
Cambridge: Cambridge University Press, 80–115.
Wodak, Ruth & Gertraud Benke. 1997. “Gender as a sociolinguistic variable: New perspec-
tives on variation studies.” In The handbook of sociolinguistics, ed. Florian Coulmas.
Oxford: Blackwell, 127–150.
Zubin, David & Klaus-Michael Köpcke. 1984. “Affect classification in the German gender
system.” Lingua 63: 41–96.
Zubin, David & Klaus-Michael Köpcke. 1986. “Gender and folk taxonomy: The indexical
relation between grammatical and lexical classification.” In Noun classes and categoriza-
tion, ed. Colette G. Craig. Amsterdam: Benjamins, 139–181.
<TARGET "chin"
</TARGET "chin">DOCINFO
AUTHOR ""
TITLE "Editors’ note"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
chinese
Editors’ note
Since Chinese is a classifier language without grammatical gender, many of the

issues that have serious implications for the description of gender in languages
such as Russian, German or Italian, are irrelevant for Chinese: nominal classes
and their gender-variable satellites, agreement, gender-resolution, so-called
generic masculines, or morphological processes of gender-specification.
Nevertheless, gender-related asymmetries can also be identified in Chinese,
where gender has more representations of a covert nature. Considering the fact
that Chinese, with its five major dialect groups, is spoken by approximately
one-fourth of the world’s population, and that – unlike in English – gender is
only beginning to be recognized as a serious (socio)linguistic issue, we have
chosen three perspectives from which gender has recently been analysed in
Chinese. Charles Ettner provides a general overview of the various gender-
related aspects in Chinese, focussing on the graphic system, word order, and the
semantic derogation of women. Marjorie Chan takes an empirical approach to
the gendered use of sentence-final particles, based on a corpus of spoken
colloquial Cantonese, while Hong Zhang analyses Chinese proverbs as a
linguistic representation of traditional cultural expectations as well as a chal-
lenge to dominant gender ideology.
<TARGET "ett" DOCINFO
AUTHOR "Charles Ettner"
TITLE "In Mandarin, men and women are equal – or – women and men are equal?"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
chinese
In Chinese, men and women are equal

– or – women and men are equal?
Charles Ettner
Stanford University, USA
1. Introduction
2. What is Chinese language?
2.1 Standard Chinese language
2.2 Structural properties of Chinese
2.3 The Chinese writing system
3. Aspects of gender in Chinese
3.1 The ‘woman’ signific and its connotations
3.2 Nouns and pronouns
3.3 Gender specification and gender abstraction
3.4 Gendered word order in Chinese
4. Social attitudes reflected in Chinese language
5. Chinese language reform and contemporary gender politics
5.1 Social language reform in a socialist China
5.2 Reproaching sexist language
5.3 China’s socialist methods for achieving reform
5.4 Promoting literacy – a double-edged sword for women
5.5 Language reform and the Script Reform Committee
6. Conclusion
Notes
References
1. Introduction
The Chinese language (Putonghua) has more than one billion native speakers,
more than any other language in the world, and roughly fifty percent of that
30 Charles Ettner
number are women. It is, therefore, unfortunate that little has been written on
gender in Chinese language. As a partial remedy, in this chapter several infor-
mative examples will be discussed which reflect extant male-dominant social
attitudes embedded in Chinese idiomatic phrases and terms of address and
reference, the Chinese writing system (graph/character composition), and Chinese
syntax. Herein, the language phenomena discussed are situated in the sociopoliti-
cal context of contemporary mainland China’s fifty years of socialist history, its
associated language reforms and language planning projects, emphasizing
successes and failures with regard to the socialist task of leveling the social
position or social status of Chinese women as compared to men. An overview
of certain relevant features of Chinese language precedes the topic discussions.
2. What is Chinese language?
“Chinese language”, in the political sense, usually refers to the seven or eight
major dialects that constitute the native spoken language of China’s majority
‘Han ethnic group’ .1 However, the term “Chinese language” challenges
linguistic sensibilities in that its major spoken dialects are mutually unintelligi-
ble. Even within many of the major dialects, there exist numerous sub-dialects
which may also be mutually unintelligible. Thus, claims of a unified Chinese
language are based on political and ethnocentric rather than on linguistic
grounds. In contrast with the spoken forms of Chinese, the writing system for
representing all Chinese dialects is essentially the same and mutually intelligible.
Dialect-specific writing systems do exist, though they are rarely used (Norman
1988: 3), and there may exist other differences. Such particularities, however, lie
beyond the scope of this discussion.
Geographically, Chinese language communities were once confined to the
Yellow river region of northern China. Over the last three millennia, the
language and its native speakers have spread across the whole of present day
China’s mainland (approx. 1 billion speakers) and Taiwan (22 million speak-
ers), and there are now expatriate Chinese communities (in total 55 million
speakers) existing in virtually every part of the globe.
2.1 Standard Chinese language

The concept of a “‘Standard Chinese language” is very much a recent develop-
ment that gained import only this century (Norman 1988: 133ff). The standard
Chinese 31
or national language of mainland China (The People’s Republic of China),

officially sanctioned in 1955, is known as pǔtōnghuà. Defined as “the
common language of China, based on the northern dialects, with the Peking
phonological system as its norm of pronunciation”, it is virtually identical to its
1930s guóyǔ ‘national language’ predecessor – the official designation of
the Republic of China government, presently in Taiwan. In English, the term
“Mandarin Chinese” serves to identify both pǔtōnghuà and guóyǔ, as linguistic
differences between the two are negligible. The focal dialect in the present
article is Mandarin Chinese (hereafter, “Chinese”). However, with the excep-
tion of pronunciation differences and a certain few lexical items, most of the
gender-related discussions have relevance to all dialects of Chinese.
2.2 Structural properties of Chinese

All dialects of Chinese are tonal and range from Mandarin, which has four
tones, to the Yuè dialect group (often labeled “Cantonese”), the members
of which can possess as many as eight or nine tones.2 Tones are phonemic, in
that they distinguish meaning among like syllables. Each tone has a fixed pitch
pattern and should not be confused with intonation or stress. Intonation also
occurs in Chinese, but its impact on tones is limited. Conversely, unstressed
syllables, especially in Chinese disyllabic words, usually drop their tone.
Chinese is essentially a monosyllabic language in which “almost every
syllable can be analyzed as an independent morpheme” (Norman 1988:154).
However, the lexicon of modern Chinese is largely polysyllabic. Polysyllabic
words have long existed in the Chinese language and are gradually increasing in
number, especially as new technologies bring about the need for new and more
complex terminologies. Chinese is an aspect language in which completion of an
action or the perfective aspect is identified by the verbal suffix -le (also called
an aspect particle).3 Tense, however, is not marked in Chinese, but is understood
either by the context of the utterance or by a specific expression of time such
as xiànzài ‘today’, míngtiān ‘tomorrow’, xiànzxài ‘now’, or
yı̌hòu ‘later’, etc. Chinese possesses a rich system of some 200 or more “measure
words” or numeral classifiers (cl) for indicating a specific number of items as
in ‘one person’ yí gè rén (one cl person), ‘three books’ sān běn
shū (three cl book), or ‘six horses’ liù pı̌ mǎ (six cl horse). Lexical
gender exists in many Chinese terms of address and reference, terms for males
and females, and in certain pronominal forms. However, gender is not marked
as a grammatical category, as Chinese is not a gender language.
32 Charles Ettner
2.3 The Chinese writing system

The earliest Chinese graphs were primarily pictographic forms, but they
gradually developed ideographic and logographic forms to include various
combinations, all of which are traditionally divided into six classes.4 Chinese
graphs are predominantly morpho-syllabic, that is, each graph represents one
morpheme which is at the same time a single syllable. However, modern
standard Chinese does possess a relatively small set of the lexicon which utilizes
two graphs to represent one syllable, such as in the word huàr ‘picture’,
which contains the graphs huà and er.
One of the most productive groups of script forms that have developed is
phonetic compounds. Each phonetic compound is composed of two elements
(e.g. mā ‘mother’): one ( ) ‘woman/female’ which is known as the “sig-
nific” (sometimes called the “radical”), and the other ( /ma/) which is
referred to as the “phonetic” component. Furthermore, each of the two
respective elements may, in itself, be composed of one or more individual
graphs (e.g. yu¹ ‘language’, which has the signific and the phonetic /wu/,
the latter consisting of the two graphs and ). The signific tends to be
located in the leftmost half of the compound, though this is not always the case.
The signific element may and often does contribute semantic meaning to the
compound as the element ‘woman’ in the ‘mother’ graph, but this is not a hard
and fast rule. The phonetic component usually contributes sound value
indicating the pronunciation, to some degree, of the medial and possibly final
segments of the word/compound graph.5
The system of graphs used in the Chinese written language differs from
that of the Latin or Cyrillic alphabet and of other similar writing systems
primarily in the specific “level of representation” of the script units. While
alphabets are phonographic systems, Chinese writing is logographic, i.e. refers
to the morphemic level (cf. Sampson 1985). Kratochvil (1968: 153f) explains
this phenomenon as follows:
While in the case of the Latin alphabet the language units which script units
were ultimately assigned to were more or less phonemes, the units of the
Chinese writing system were assigned to language units of, so to speak, a higher
level. The two systems thus differ in nothing less and nothing more than the
level of representation of language units by script units. […] Chinese script is
pictographic by its origin, but it is not pictographic in its present function,
since characters are not used as pictures of objects and acts but as ‘pictures’ of
certain language units, and it is not ideographic, since characters are not direct
symbols of ideas or abstract concepts but symbols of certain language units
Chinese 33
(which in their turn may perhaps be considered as symbols of ideas or abstract

concepts but this is not a directly relevant consideration). […] the units of the
Chinese writing system represent certain language units which are partly
characterized by specific features of meaning […]
A particularly important point of Kratochvil’s discussion is the fact that

graph/character “units of the Chinese writing system” continue to be “partly
characterized by specific features of meaning”. The further significance of these
linguistic features as it pertains to gender and the relevant social relationships
they inform is elaborated specifically in the discussion that follows. It should
also be noted that the romanizing system for Chinese characters most popular
today is mainland China’s “Pinyin” system, which is also used in this chapter.
3. Aspects of gender in Chinese
The lexical and structural properties of Chinese language display evidence of

ancient – and, I suggest, still extant – social attitudes which exhibit a distinct
relationship of language and gender. Such evidence exists particularly among
groups of Chinese graphs (characters), pronominal word forms, and in the
specific word order constraints for separate terms denoting female and male in
the context of a noun phrase. Notably, certain linguistic remnants shed light on
a long forgotten China antedating all known Chinese written history.
Chinese orthography (the manner in which graphs are constructed)
provides interesting clues about the nature of ancient Chinese social structure,
and about attitudes and values concerning the status of women at various
historical periods. Evidence exists to suggest that during the earliest formative
period of the Chinese written language, ancient China may have been a matri-
lineal society. Remnants of the social character of ancient Chinese are pre-
served in particular words like shí ‘ancestor’ and xìng ‘surname’, both of
which contain the signific (the element which often provides semantic mean-
ing) nu¹ ‘female’. That words referring to family origins and familial ties
should contain an element specifically denoting female can hardly be seen as
insignificant, coincidental, or arbitrary. Similarly, a number of the more ancient
surnames: Yáo, Jı̄, Jūng, and Lóu, for example, also incorporate the
nu¹ signific within the graph (Tang 1988: 61). This sign and symbol relationship
is also exemplified by other graph groups within the Chinese language.6
34 Charles Ettner
3.1 The ‘woman’ signific and its connotations

Chinese compound graphs contain a signific component which often contrib-
utes meaning to the lexical item it helps to form. Tang surveyed the Chinese
lexicon for words containing the signific nu¹ ‘female/woman’. The results of
the survey were divided into four semantic categories: (1) Words relating to
marriage or giving birth, (2) kinship terms and terms regarding family relation-
ships, (3) words referring to beauty, and (4) derogatory words or words with
negative connotations (Tang 1988: 62).
In a survey I conducted using a modern Chinese dictionary, of the more
than 500 entries comprised of compound graphs with the nu¹ ‘female/
woman’ signific, approximately 20% were of the above fourth category,
“derogatory words or words with negative connotations”.7 Subsequently, I
made additional surveys of earlier published and other contemporary Chinese
language dictionaries, and all yielded similar results.8 Unfortunately, it was not
as easy to contrast the Chinese graph for ‘man/male’ nán in my surveys due
to the fact that is, in itself, not a signific. Rather, is but a compound graph
comprised of the tián ‘field’ and lì ‘strength’ graphs. Furthermore, of the
twenty-odd listed compound graphs with a nán graph component, only one
‘angry’ communicates a derogatory or negative meaning. However, the other
component of that one graph is the nu¹ female signific! Going one step
further, of the forty-seven Chinese compound words containing a nán graph
component, only two listings, dànánzı̌zhǔyì ‘male chauvinism’
and jìnán ‘male prostitute’, have negative connotations. In the latter of
these two, once again, one finds a compound graph with the nu¹ female
signific. Nonetheless, it is important to also note that, isolated as a single
graph, nu¹ connotes neither positive nor negative meaning.
In some Chinese compound graphs, meaning is contributed not only by the
signific, but also by the phonetic element. An example can be found in the
word nù in which the signific xı̄n ‘heart, mind, feeling’ or ‘intention’ is
combined with the phonetic nú ‘slave’ or ‘enslave’ to yield the meaning of
‘anger, rage’, or ‘fury’. Again, there exists a significant number of this type of
compound graphs containing nu¹ as part of the phonetic component and
connoting negativity. And although the nu¹ graph combines variously with
other graphs to form different phonetic elements, the majority of these phonetic
elements are not homonyms of /nü/ and, therefore, are only visibly recognizable
as implicitly related to female/woman. There are also numerous related
compound words of the same type.
Chinese 35
A short list of example words containing the nu¹ signific element that I
identified in the survey as “derogatory” or “with negative connotations” are
listed below.
Table 1.Graphs containing the signific element
Graph Pinyin Meaning Graph Pinyin Meaning
chı̄ ugly màn scorn, humiliate

dù jealous, envy nú slave, enslave
fāng harm pı̄n to have illicit relations
fáng hinder, hamper, impede xian suspicion, ill will, dislike
jí jealous, hate yāo force, coerce
jiān wicked, evil, treacherous yāo goblin, demon
The above table9 displays lexical items consisting of single morphemes and
single “compound graph” forms. However, there also exist compound word
forms in Chinese language, that is, lexical items consisting of two or more
morphemes and hence two or more graphs. The significance of this fact is that
the effect of each negative root is multiplied by the number of compound word
forms it may yield. Examples of compound word forms of this nature can be
seen in Table 2:
Table 2.Compound word forms containing two graphs
Graph Pinyin Meaning
jiānchén treacherous court official

jiānxiǎn wicked and crafty, treacherous, malicious
jiānguı̌ evildoers, malefactors
jiānxiào sinister (villainous) smile
jiānhuá treacherous, crafty, deceitful
jiānxié craft and evil, treacherous
jiānnìng crafty and fawning
jiānxióng a person who achieves high position by unscrupulous scheming
jiānshāng unscrupulous merchant, profiteer
jiānyín illicit sexual relations; adultery, rape
jiānwū rape or seduce
jiānzéi traitor, conspirator
jiānxi spy, enemy agent
jiānzhà fraudulent, crafty, treacherous
Summarizing the discussion thus far, there is ample evidence to suggest that a
visually apparent connection exists in the graph forms of the Chinese writing
36 Charles Ettner
system between the representation of females/women and various negative or

derogatory concepts. The evidence presented earlier also suggests that this
connection is very likely a derived representation of ancient Chinese societal
conceptions and attitudes about women. Furthermore, there is little, if any,
evidence to suggest that contemporary Chinese society has effectively or ade-
quately divorced itself from these same traditional conceptions and attitudes.
3.2 Nouns and pronouns

Contemporary Chinese language displays a gender distinction only in second
and third person pronoun forms. There also exist the lexical gender nouns
nán ‘male’ and nu¹ ‘female’ for humans, gōng ‘male’ and mǔ
‘female’ for animals, and a generic form rén ‘person’ or ‘people’. The
pronominals nı̌ ‘you’ and tā ‘he’ or ‘she’ were historically devoid of
gender distinctions until recent times.
Around the late 1910s writers began to use three separate graphs to distinguish
‘he’, ‘she’, and ‘it’ [ , and , respectively]. All three characters were
pronounced exactly the same way [tā][…] But the use of separate graphs gave
the illusion that Chinese had the same kind of gender distinction that many
Western languages did. (Ramsey 1987: 50)
Nonetheless, it is still true that when oral communication is non-specific or

uncertain with regard to gender, native speakers will normally assume the pro-
nominal referent of tā to be a male person. Note also that the ‘male’ pronoun is
identical with the undifferentiated pronoun and is written with the ‘person’
signific. Similarly, the supposed gender-neutral term ‘person’ or ‘people’ rén
might better be characterized as a male generic term. In other words, nı̌, tā, and
rén have a supposed generic quality in oral contexts (and historically, in written
contexts as well), but each connote the male gender by default.
The second person pronoun nı̌ ‘you’ also acquired a complementary
female graph form sometime during this century, although it has yet to enjoy
significant usage among the general populace. Curiously, the second person
polite pronoun nín ‘you’ still has but a solitary, and supposedly, gender-
neutral graph form.
3.3 Gender specification and gender abstraction

As noted above, the inherent nature of the pronominals tā ‘s/he’ and tāmen ‘they’
is such that each can produce gender ambiguity in spoken communication.
Chinese 37
Moreover, they are the primary resource for achieving intentional gender
abstraction, in other words, for neutralizing gender in spoken Chinese. On the
other hand, tā and tāmen each have the same level of specificity (or lack thereof)
as their subject or object referents; cf. examples (1) and (2):
(1)
Wáng tàitai shì wǒ-de laǒshı̄. Tā shì Zhōng.guó rén.
Wang Mrs be I/me-poss teacher she be China person
‘Mrs. Wang is my teacher.’ ‘She is Chinese.’
(2)
Xuésheng dōu kăo wàn le. Tā-men kuài yào xiàkè.
student all test finish asp s/he-pl quick be.going.to finish.class
‘The students all finished the test.’ ‘Their class will soon be over.’
Often nouns can be gender-marked simply by preposing either nu¹ ‘female’

or nán ‘male’ to the noun, thus, effectually yielding a gendered compound
noun; cf. examples in (3):
(3)
rén nu¹ -rén nán-rén
person female-person male-person
‘person’ ‘woman’ ‘man’
xué-sheng nu¹ -sheng nán-sheng

study-scholar female-scholar male-scholar
‘student’ ‘female student’ ‘male student’
háizi nu¹ -háizi nán-háizi

child female-child male-child
‘child’ ‘girl’ ‘boy’
As Chinese is not a gender language, structurally and historically, nouns and

pronouns are inherently ambiguous with respect to gender. Only by preposing
lexical gender words to such word classes, either as noun modifiers or pronom-
inal antecedents, does gender specificity obtain.
3.4 Gendered word order in Chinese

Chinese syntax provides an interesting illustration of the archaic hierarchical
attitudes that have become embedded in word order constraints for Chinese
38 Charles Ettner
spoken and written language. This phenomenon exists as a grammatical rule

which, in essence, requires that gender-specific words be ordered within noun
phrases (NP) so that the male-gendered constituent precedes the female-
gendered constituent. To reverse the prescribed order and place female ahead of
male renders the sentence ungrammatical. Consider the sample sentences in (4):
(4) a.
nán nu¹ píng.děng
man woman equal
‘Men and women are equal.’
b. *
nu¹ nán píng.děng
woman man equal
*‘Women and men are equal.’
Chinese graphs may be written vertically, from top to bottom, or horizontally,

from right to left or left to right (but never from bottom to top). However, the
male-precedes-female word ordering is required regardless of the direction in
which the graphs are written. Furthermore, the same gender ordering con-
straints obtain for several gender-specific words/graphs pertaining to the animal
kingdom such as gōngmǔ ‘male and female’ (of animals), fènghuáng
‘phoenix’ (male and female), and yuānyang ‘Mandarin ducks’ (male and
female).10 Still, the rule as I have formulated it is somewhat oversimplified, as
the following example illustrates:
(5)
wǒ jiā yǒu bàba māma jiějie mèimei
my family have father mother oldr.sister yngr.sister
dìdi hé wǒ
yngr.brother and me
‘In my family, there is my father, mother, older sister, younger sister,
younger brother, and me.’
Example (5) reveals a more fundamental hierarchy of NP constituent ordering

which, by no mere coincidence, mirrors the essence of the Confucian social
order of the propriety of relationships (i.e., tiāndi ‘heaven and earth’,
ju¶ n-chén ‘sovereign and subject’, fù-zı̌ ‘father and son’, fūqı̄
‘husband and wife’, and xiōngdì ‘older brother and younger brother’).11
This propriety of social relationships reduces to three ordered levels of hierarchy
Chinese 39
which permeate much, if not most, of Chinese social culture, including Chinese
language. The three ordered levels are: first, social position (e.g., sovereign
before subject); second, generation/age (e.g., father before son); third, gender
(e.g., husband before wife).12 This same ordering of relationships exists in the
syntactic phrase structure of NP constituents in Chinese language: The noun
constituents in an NP proceed from highest to lowest in hierarchical order.
Hierarchical ordering proceeds according to ranked categories: (1) highest
social rank or position; (2) same-gender siblings; (3) eldest generation/age; (4)
gender (male precedes female).
The above grammar accounts for the three categories of social order and an
additional “same-gender siblings” category that is not necessarily part of the
prescribed Confucian social order, but is very much an element of the tradition-
al and extant Chinese social order. Table 3 below illustrates the grammatical
constraints of Chinese NP constituent ordering using an optimality theory
schematic (cf. Prince & Smolensky 1993).13
A notable exception to the above NP constraint rule is the term
yı̄nyáng (essentially, ‘female-male’). While it may be possible that the extreme
antiquity of the term antedates Chinese male social hierarchy, Chan (1998)
offers evidence of competing phonological constraints involving a “preferred
prosodic sequence” of NP constituents “with respect to the historical four-tone
categories of Ping, Shang, Qu and Ru ( )” and voiceless and voiced
initial phonemes (Chan 1998: 46; cf. Ting 1969, Lien 1989). In some instances
both prosody and semantics converge in supporting the gender-sensitive
arrangement of NP constituents. However, the additional evidence of ‘voiceless
precedes voiced initial’ is particularly useful in accounting for rare exceptions
such as yı̄nyang, and two instances of animal terms pìn mǔ ‘cow and
bull’ and cíxióng ‘hen and rooster’, all of which had voiceless and voiced
initials, respectively, in Middle Chinese (i.e. pre-modern Chinese): “In coordi-
nate compounds […] the syllable with a voiceless (qing ) initial typically
precedes that with a voiced (zhuo ) initial” (Ting 1969). The sequence of the
constituents in yı̄nyáng “ ‘yin and yang’, a historically well-established
compound, obeys this phonological constraint of ‘qing-before-zhuo’, and not
the semantic one of ‘male-before-female’” (Chan 1998: 46; cf. also Lien 1989).
40 Charles Ettner
Table 3.Word order constraints in Chinese NPs

Order of noun Rank Gender pairs Generation Gender
constituents superior>> same gender>> elder gen >> male >>
subordinate not same younger gen female
– king – peasant
peasant – king *
– father – son
son – father * *
– husband – wife
wife – husband ? *
– elder bro – yngr bro
yngr bro – elder bro *
– elder bro – elder sis
elder sis – elder bro *
– elder sis – yngr bro *
yngr bro – elder sis *
– eld&yngr bro – elder sis *
elder bro&sis – yngr bro * *
– eld&yngr sis – yngr bro *
elder sis – yngr bro&sis * *

(–) indicates preferred order
(*) indicates category conflict
4. Social attitudes reflected in Chinese language14
That remnants of the social character of ancient Chinese society are preserved
in certain features of contemporary Chinese language is not surprising. Such is
likewise true of many other languages throughout the world. How such
remnants – notably, male-dominant, Confucian social attitudes – came to be
embedded in Chinese language may be explained within the framework of
Chinese 41
linguistic relativity as suggested by Sapir (1933: 17), who stated that “[t]he
language habits of people are by no means irrelevant as unconscious indicators
of the more important traits of their [social] personalities […]”. Sapir makes
the further point that these are embedded in a variety of linguistic forms which
endure through time:
A great deal of the cultural stock in trade of a […] society is presented in a
more or less well defined linguistic form. Proverbs, medicine formulae,
standardized prayers, folk tales, standardized speeches, song texts, genealogies
are some of the more overt forms which language takes as a culture-preserving
instrument. (Sapir 1933: 16f)
For Chinese language, this “cultural stock” took the form of the dominant
social attitudes and values that were embodied in Confucian ideology, and
which embraced a hegemonic, patriarchal social system.
Stacey (1983: 30) points out that “[f]ew patriarchal systems approach the
degree of explication, elaboration of hegemony Confucianism achieved in
imperial China”. Moreover, this particular system applied not only to common-
ers, but to the emperor and all his subjects, and it became the model for both
the social and political hierarchy in China.
In essence, Confucianism was a protocol for proper family life. It prescribed a
patriarchal, patrilineal, and patrilocal family system. Sex, age, and generation
were the coordinates that defined an individual’s status, role privileges, duties,
and liabilities within the family order. Men officially dominated women; the
old dominated the young. (Stacey 1983: 31)
In codifying the Confucian system, it was incorporated into the ‘code of ethics’
lı̌jiào, and eventually became part of Chinese jurisprudence. “Li and law
were the major vehicles through which Confucian doctrines were promulgated
and transmitted” (Stacey 1983: 35). Over time, Confucian doctrine became
embedded in virtually every aspect of Chinese society, including language, as
the force of these social attitudes became more deeply enveloped in Chinese
law. “The law served to codify and reinforce the basic authority principles of
Confucian patriarchy. Patriarchal authority was officially recognized, strength-
ened, and occasionally implemented by the dynastic legal codes” (Stacey
1983: 37). Not until the advent of socialist doctrine and the newly established
People’s Republic of China (PRC) government was Confucian doctrine ever
seriously challenged.
Not unrelated to patriarchy and male dominance, a certain disparate
valuation between Chinese males and females is often reflected in the tradition-
42 Charles Ettner
al gender-stereotyped ways of naming children as well as in the terms of address

and reference for ‘husband’ and ‘wife’. Most often, the naming of females
accords with the notion that women are expected to be “virtuous”, “beautiful”,
and like “nature”. Therefore, females often receive names such as shū ‘good
virtue’, zhēn ‘chastity’, shèn ‘prudence’, shùn ‘obey’, hui ‘kindness’,
měi ‘beauty’, yǎ ‘elegant’, yuán ‘beautiful girl’, wǎn ‘gentle’, xián
‘refined’ or ‘demure’, lán ‘orchid’, jú ‘chrysanthemum’, fēng ‘wind’,
huā ‘flower’, xuě ‘snow’, yuè ‘moon’, jı̄n ‘gold’, yín ‘silver’, bǎo
‘treasure’ or ‘precious’, yù ‘jade’, and zhēn ‘treasure’ (Tang 1988: 61).
By contrast, males usually receive names like chāng ‘prosperous’, róng
‘glorify’, hūi ‘glorious’, huáng ‘bright’, hóng ‘great’, yào ‘bright
light’, xióng ‘heroic’, kúi ‘giant’, jié ‘heroic’, fàn ‘exemplify’, zōng
‘respect ancestry’, and zǔ ‘ancestry’. In all, a male’s given name serves to
signify the hope that he will “glorify the family or become prosperous” as is his
expected role in Chinese society (Tang 1988: 61f). Conversely, there are few
such glorious expectations attached to the lives of Chinese females.
Chinese terms used in addressing or referring to one’s wife or husband even
more vividly denote the social positions and relative values traditional Chinese
society assigned to males and females, respectively. Historically, a female spouse
referred to herself as nújiā (literally) ‘house slave’ or qiè ‘concubine’,
while, on the other hand, she referred to and addressed her husband as
dāngjiāde which denotes ‘master of the family’. There also exists a host of terms
that simply translate as ‘husband’, but many of these terms also connote
respect and/or authority as well (Tang 1988: 62). Consider the following
examples: xiānsheng ‘gentleman, husband’ or ‘teacher’, zhàngfu ‘hus-
band’ or ‘manliness’, and liángrén ‘good person’ or ‘husband’.
Among the more common terms of address and reference a husband might
use regarding his wife are (apart from terms strictly denoting ‘wife’) nèirén
‘inside (the house) person’, jiànnèi (literally) ‘inferior inside’, nè zhù
‘inside help’, huáng liǎn pó ‘yellow faced woman’ (‘yellow faced’ implies
an old and, thus, ugly face), chángshé fū ‘long tongue woman’ (implies
a gossip), zhū fān de ‘rice cooker’ or ‘a cook’, and háizi de mā
‘mother of the children/the children’s mother’ (Tang 1988:62). These and all of the
foregoing examples well illustrate the fact that belittling and demeaning terms of
address and reference in Chinese language predominantly concern females,
while respectful and honorific terminologies conversely pertain to males.
Chinese 43
5. Chinese language reform and contemporary gender politics
Leaving linguistics for the moment, the specific discussion of gender issues
concerning Chinese women has been increasingly gaining a public voice and
public attention throughout the Chinese speaking world. In mainland China,
the Chinese Communist Party (CCP) affiliated “All-China Women’s Federa-
tion” has, for some fifty years, been responsible for much, if not practically all,
of the country’s research and problem solving efforts regarding women and
women’s issues (Jacka 1997: 7). The Women’s Federation, with some 90,000
professional cadre and 56,879 branches, is China’s largest non-governmental
organization (Hall 1997: 173). However, its existence has always been depen-
dent on a favorable rapport with the CCP. The federation’s various publica-
tions, the magazine New Women of China, the monthly periodical Chinese
Women, the newspaper Chinese Women’s Daily, and the internal periodical
Woman-work serve to inform and educate the Chinese public about required
reforms and of recent progress toward such ends. However, the Women’s
Federation has continually struggled with its “dual task […] to mobilize mass
support of CCP policies, and to defend and further the interests of women”.
This real conflict, which both the Women’s Federation and the CCP leadership
deny, has “sometimes resulted in an acceptance by the Women’s Federation of
Party policies that are in fact detrimental to women’s interest, and a failure, on
its part, to tackle problems for women that the Party itself does not acknowl-
edge, or regards as unimportant” (Hall 1997: 30ff).15
Gender politics in mainland China have largely focused on issues of
inequality in the workplace, in hiring practices, in access to education, in
marriage and divorce practices, and in land ownership laws. Seldom have they
engaged the linguistic realm such as language planning policies or language
change, and only fleetingly has attention been given to sexism in Chinese
common speech or within Chinese language specifically. Not surprisingly, labor
and other subsistence-related issues continue to draw a much higher political
priority in China, at both the individual and official levels, than do most
concerns for language policy and other language reforms. Notable exceptions
include the Chinese script or character simplification project and certain
socialist-flavored language reforms of the 1950’s. The latter group of socialist-
inspired reforms focused upon specific terminologies, terms of address, and
idiomatic phrases regarding women.
The state of contemporary gender politics in most other Chinese communi-
ties, such as in Singapore, Taiwan, and in the now repatriated Hong Kong
44 Charles Ettner
community appears similar to that of mainland China. Taiwan, for example,

has numerous feminist groups that actively campaign for reforms, and with
little fear of coercion or interference from the Taiwanese government. Yet, the
bulk of feminist reform projects and popular feminist issues publicized in
Taiwan mirrors many of the subject issues of its mainland Chinese neighbor.
Again, labor and other subsistence-related issues list among the most important
of Taiwanese feminist concerns while, again, attention to Chinese language
reform, language planning policy, and other areas of linguistic concern have
been slight, if at all existing.16
5.1 Social language reform in a socialist China

The new PRC government set out almost immediately to reorder Chinese
society and to establish a more egalitarian environment by ordering a massive
redistribution of all Chinese land holdings and by legislating changes such as a
marriage reform law designed, in part, to improve women’s rights and to level
their social status with men. The land reform of 1950–1953 brought with it not
only the redistribution of some 47 million hectares of farmland among 300
million peasants (Gao 1987: 374), but a new social political awareness concern-
ing class, social status, education, and socialist ideologies. The 1950 Constitu-
tion asserted the equality of all Chinese citizens, including women and all
Chinese Minority Nationalities: “The marriage law of the same year gave
women the right to choose their own marriage partners and to demand a
divorce. Women’s feet were no longer bound, even in remote villages; girls were
given at least limited access to education, and more” (Honig & Hershatter
1988: 3; see also Stacey 1983: 176, 181). However,
[d]irect campaigns to implement the marriage law came to an end in 1953
mainly because the fundamental goals of family policy of the new democracy
had been achieved. These always had subordinated women’s liberation to the
priorities of the class struggle. (Stacey 1983: 181f, emphasis added)
The CCP was steadfast in its belief that “class struggle” was the methodology by
which the Chinese people’s socialist, egalitarian reforms would be realized,
presumably effecting women’s “liberation”. As a product of these socialist
reforms, “new terms and expressions necessitated or motivated under new
political and social conditions” were increasingly added to public vernacular to
explain and reinforce the PRC’s Marxist/socialist project (Tai 1975: 234).
Yet, even before the founding of the People’s Republic of China, the
Chinese 45
Chinese Communist Party had adopted various terms of reference and address
essentially devoid of either social status or gender-specific connotations.
Probably, the best known of these is the socialist term tóngzhì ‘comrade’,
which is added post-positionally to one’s full or surname, as in ‘Com-
rade Wang’. Tóngzhì had originally meant ‘having the same will or interest’. In
the early period of the PRC, from 1949–1965, Chinese people resisted replacing
honorific titles with newly re-defined terms. But, from the start of the Great
Proletariat Cultural Revolution (GPCR) onward, the use of tóngzhì gained
popularity. Towards the end of the GPCR, on June 2, 1975, the Rénmín Rìbào
(People’s Daily) carried an article admonishing the Chinese people to refrain
from using terms of rank or position and to revitalize their use of tóngzhì. The
impetus for the article came from the concern of the Party Central Committee
and the State Council over the increasing re-emergence of hierarchical terms of
address (Fang & Heng 1983: 499). In 1976, after the defeat of the Gang of Four,
official and professional positions were re-established, and their original titles
regained use. However, in 1977, the government reaffirmed its former position
in favor of tóngzhì, an attitude it has maintained ever since (Scotton & Wanjin
1983: 479f).
In recent years, use of the term tóngzhì, once commonplace in mainland
China, has seen a substantial decline, except in highly public and official
forums. The same holds true for most of the many redefined socialist terminol-
ogies. Conversely, one of the most successful socialist-born terms of address is
that of àirén, used to denote ‘husband’ or ‘wife’. Ironically, the original
meaning of àirén is ‘lover’, and that remains unchanged for Chinese speakers
outside of mainland China. Not surprisingly, the rest of the Chinese-speaking
world has shown little interest in adopting any of China’s various redefined
socialist terms of address or terms of reference.
5.2 Reproaching sexist language

The 1950s onslaught of socialist projects provided ample impetus for the
emergence of an offensive against a number of Chinese words, titles, and labels
denoting non-socialist hierarchy or stratification, as well as against sexist
language. As a result, socialist replacement terms came into being, while the
words they replaced, along with a host of sexist idioms, were summarily
abolished from public and, allegedly, private use. The following examples typify
the general character of the sexist idioms and proverbs that the PRC’s socialist
reforms targeted:
46 Charles Ettner
(6)
cóng yı̄ ér zhōng
follow one until *finish
‘A woman marries one husband to the end (till her death).’
(*) meaning derived from context17
(7)
jià jı̄ suí jı̄
marry rooster follow rooster
‘Whatever husband a woman marries, she must follow him her whole life.’
(8)
sān cóng sì dé
three follow four virtue
‘A woman must follow three people (her father, husband and son) and
four virtues (faithfulness, filial piety, chastity, and kindness).’
(9)
sān gè nǔ rén luàn tiān xià
three cl female person chaos heaven below
‘Three women can bring the whole world into chaos.’
(10)
shí gè nǔ rén jiǔ gè dù
ten cl female person nine cl jealous
‘Nine out of ten women are jealous.’
(11)
wéi nu¹ .zı̌ yǔ xiăo rén wéi nán yăng yě
only woman and small person be difficult support also
‘Women and uneducated people are the most difficult to deal with.’
Language reforms concerning spoken language lacked formal names or

campaigns, but were undeniably a part of the socialist project. Particular
attention was given by cadre, Party officials, the press, and by other believers in
the socialist project to eliminate any language form that conflicted with socialist
ideology both in the whole of the period preceding, and even more so during
the GPCR. One of the areas of focus concerned non-egalitarian language:
Egalitarianism is one of the most emphasized aspects of the Chinese commu-
nist ideology. As a result, most expressions making hierarchical distinctions
have been dropped […] Egalitarianism has also caused the abolition of
expressions praising exploiting and bourgeois classes, expressions contemptu-
ous of the working class […] Similarly, the egalitarianism has made obsolete
Chinese 47
expressions discriminating against women, juniors, and minority nationalities.

For example, nuziwucai bianshi de (a woman is virtuous when
she is incapable), nianshao wuzhi (ignorant because of young age),
and nanman (southern uncivilized races) […] While terms implying
traditional hierarchical distinctions have been demoted, terms of class distinc-
tions based on Marx’s theory of class struggle have been promoted.
(Tai 1975: 236f; cf. also Shih 1984, Farris 1988, and Tan 1990)
5.3 China’s socialist methods for achieving reform

Changing the speech habits, even more, the worldview of an entire country –
and particularly of a nation the size of China – is by no means a simple task.
Moreover, these specific reforms were not the well-orchestrated project of any
formalized Chinese language planning efforts. Rather, they derived from the
essentially non-linguistic political project of socialist reformation. Nonetheless,
gaining compliance with these as with most of the PRC’s socialist reform efforts
was, perhaps, the government’s least area of difficulty. The primary reason for
this is that compliance through coercion was then, and has long been, typical of
the PRC’s numerous social and political campaigns, the most well-known being
the infamous Great Proletariat Cultural Revolution.
The Chinese people’s acceptance of a new Chinese lexicon of political and
socialist terms and slogans was a direct result of “the deliberate manipulation of
the [Chinese] language as a tool to achieve socio-political reforms, and [to
stimulate] the emergence of new social values and new modes of thought” (Tai
1975: 239f). Public confession meetings inspired the acceptance of the egalitari-
an reforms and the abolition of non-egalitarian language. “People were encour-
aged [under duress] to confess publicly errors of ‘feudal’ thought” (Stacey
1983: 180). Note that while “feudal thought” (and language derived from feudal
thought) was a genuine concern of the socialist movement, it was equally a
convenient candidate for the frequent exaggerations of coerced, public “struggle
meetings”.18 Thus, one can reasonably deduce that the social decline of sexist
idioms in Chinese speech may, at least in part, be due to the deplorable and
reckless nature of the struggle meetings themselves.
Unfortunately, this tool of coercion, which was used to stimulate new
modes of thought and compliance for the benefit of Chinese women, later
became a double-edged sword that was used to suppress them:
After the 1950 Constitution guaranteed equality between men and women, and
after sweeping attacks were made in the 1950’s on the abuses and injustices
48 Charles Ettner
suffered by women in the “old society”, Chinese government publications

generally asserted that the task of liberating women had been accomplished.
(Honig & Hershatter 1988: 309)
With the political tables turned, the impetus for further feminist reform,
including language reform, was not only absent but contrary to the party line.
“[W]omen were primarily mobilized not to fight for gender equality, but to
contribute to socialist construction. It was widely accepted that the establish-
ment of socialism would automatically result in the liberation of women”
(Honig & Hershatter 1988: 3).
In the years that followed, the political campaigns grew more intense. By
the onset of the GPCR in 1966, the political thrust of “class struggle took
precedence over all other issues, including those of gender equality […] In fact,
recognition of gender as a category distinct from class was regarded as reaction-
ary” (Honig & Hershatter 1988: 3f). The most interesting aspect, in regard to
these statements, is not that gender equality issues and their related reforms
were no longer touted as socialist concerns, but that the PRC government has
yet, following the end of the GPCR period, to enthusiastically support or
reassert the importance of gender equality in China as once again constituting
a substantive socialist concern. Conversely, the government did reassert
concern for maintaining the use of socialist-inspired egalitarian terms of
address such as tóngzhì ‘comrade’, and other terms.
5.4 Promoting literacy – a double-edged sword for women

The PRC government, from its inception, promulgated what turned out to be
a very successful campaign to instill at least some degree of literacy amongst a
high percentage of the Chinese populace, and thereby established a vehicle for
disseminating socialist education. This same literacy vehicle, however, also
provided the Chinese people with access to the visually symbolic representa-
tions of females in written forms of the Chinese language, and likewise to the
equally obvious demeaning association of the nu¹ graph with words having
negative connotations. As Sapir (1933: 28) explains: “Not only a language or a
terminology but the mere external form in which it is written may become
important as a symbol of sentimental or social distinction.” Thus, I submit that
it would be virtually impossible to acquire literacy in Chinese without also
gaining at least a matter-of-fact realization of the symbolic, semantic association
of ‘female’ with socially negative traits or characteristics.
Chinese 49
5.5 Language reform and the Script Reform Committee

The evidence above suggests that the socialist reforms of the formative PRC
period did make some attempts to eradicate sexist idioms and establish gender-
leveling, egalitarian language amongst the Chinese people. However, these few
attempts had only limited, short-term success with most of the new terms of
address or reference, and virtually no lasting success with regard to commonly
used sexist idioms, save for in political or otherwise highly public forums. Still
other government attempts at Chinese language reform were orchestrated by
the wénzì gǎigé wěiyuánhuì ‘Script Reform Committee’ (SRC)19
which focused upon simplifying Chinese characters, spreading the use of the
standard vernacular, and devising and disseminating a phonetic alphabet for
use as a tool for learning Chinese.20
The SRC’s efforts, with regard to its three-fold task, were relatively success-
ful, though not without difficulties.21 Lehmann (1975: 45) found troubling the
rather “ad hoc nature of the main language planning agency and the exclusion
of lexicography and technical vocabulary development from its responsibili-
ties”. Seybolt & Chiang (1979: 11) also pointed out that the SRC had “no formal
principles for simplifying characters”.22 Furthermore, a first article in
‘Chinese Language’ made public these issues stressing a need for scientific
studies in lexicology, sociolinguistics, language and communication, and other
neglected areas (Cheng 1986: 89). Also excluded from the SRC’s agenda were
the various socio-political and egalitarian language reforms, which seems
particularly odd in view of statements like “The language reform which we are
now carrying out is part and parcel of our socialist revolution and construction
and serves to consolidate the dictatorship of the proletariat” (Wen 1973: 12 “On
Reforming Written Chinese” as quoted in DeFrancis 1984: 267). Similarly,
Chang Hsi-jo, the former Minister of Education, maintains that “language
reform is a serious political mission which must be carried out in order to
strengthen the solidarity of the people of the whole country” (quoted in
Tarantino 1975: 107).
This Chinese “socialist revolution” and “dictatorship of the proletariat”
was never intended to exclude Chinese females; however, the project of the
SRC effectively did so! Script reform might well have addressed reforming
nu¹ signific-type graphs connoting negativity by, perhaps, replacing it with the
dǎi ‘evil’23 or rén ‘person’ signific or with some other simplification or
design of their choosing. This might not be possible for all such graph forms,
but it surely could reduce significantly the volume of word forms of this nature.
50 Charles Ettner
Yet, the failure to address this problem should not reflect on the SRC members,
per se, but rather on the body politic directing them.
6. Conclusion
In this chapter, an attempt was made to show that certain features of contempo-
rary Chinese language reflect male-dominant social attitudes and, conversely,
notions of female inferiority still extant in Chinese society. The evidence above
suggests, for example, that Chinese terms of reference and address do exhibit
gender disparity between husbands and wives much like Chinese naming
practices for infant children illustrate the asymmetric, stereotyped expectations
attached to males and females, respectively. The PRC’s short-lived socialist
reforms aimed at abolishing sexist idiomatic phrases offer yet further evidence
of the extremely male-dominant social attitudes that still abound.
The fact that nearly 20% of Chinese words containing the nu¹ ‘female’
signific connote negativity and no similar words contain the graph nán
‘male’ suggests that female debasement also extends to visual representations in
the very orthography of Chinese language. And, in both spoken and written
Chinese, the male-precedes-female word order constraints illustrate a male-
dominant hierarchical set of rules operating on the syntactic structure of
Chinese language.
The above discussions also surveyed the character of the PRC’s socialist
reform projects and, particularly, the sociopolitical environment of its formal
and informal language reforms since 1949. An attempt has been to explicate the
interplay of these many reforms and reform projects, the degree and manner in
which they addressed or sought to redress sexism in Chinese language, and the
resulting success or failure they have achieved in promoting the equality of
Chinese women.
China’s attempts at socialist language reform and the monumental accom-
plishments of the SRC are certainly to be applauded. At the same time, the
Chinese government’s failure to coordinate and/or conjoin a reform of male-
dominant features in Chinese language and other aspects of socialist language
reform with the SRC’s formalized three-fold project seems no less worthy of
criticism. The fact that the SRC’s project tasks did not include socialist language
reform makes for a weak defense since that in itself belies the precept that
language reform is part of the Chinese “social revolution” and a “serious
political mission”. Serruys (1962: 57) explains that
Chinese 51
[f]or the Communists language reform is needed to serve the demands of mass
information and mass propaganda among the people and to further the cause
of the continuing Communist revolution and the realization of the true
Communist society.
How then could Chinese language reform “further the cause” by neglecting the
social revolution of fully one-half of “the true Communist society”, namely,
Chinese females?
The potential argument that a reform of the male-dominant features in
Chinese language would be too difficult to implement effectively also runs
counter to Chinese experience. The very effective coercive compliance tactics of
the Chinese government are documented as well as the campaigns and projects
they have been used in – this being the government’s chosen method, not my
own. In other instances, charismatic government leaders have by word alone
induced the Chinese people to take on and carry out even ultra-radical reforms
such as the ten-year Cultural Revolution. Methodologies notwithstanding,
effecting public compliance with PRC government directives historically has
not been problematic. Therefore, had the Chinese government fostered within
its language reform campaign countermeasures to the male-dominant features
in Chinese language, the evidence suggests that the campaign would have
succeeded, at least to a noticeable degree. Similarly, one can only fathom the
effect, had Mao Zedong espoused a reform campaign of the Chinese populace
directed against all forms of sexism in Chinese language, in the interest of
socialist revolution and the dictatorship of the proletariat.
Elsewhere in the world, no movement or campaign has yet effected the total
elimination of sexist language among its subject people. And yet, success in each
instance exists not merely in the resulting reduced usage of sexist language
forms, but also in the people’s newly increased awareness and heightened
sensitivity to sexist language, and especially in the irrevocable seeds of language
change and reform that were planted. For similar reasons, the above hypotheti-
cal Chinese language campaign would not have required total compliance for
it to have achieved success. Mao Zedong once stated “The day all Chinese
women stand up will be the time of victory of the Chinese Revolution” (Croll
1976: 87). However, such hypothetical victories will continue to be unrealized
so long as the requisite reforms preceding them remain likewise hypothetical.
The struggles of Chinese women to attain gender parity in the workplace,
equal pay for equal work, and equal access to jobs, to quality education, and to
the Chinese business arena are all important, ongoing projects. And yet, these
subsistence-oriented struggles seem one-sided in that, metaphorically speaking,
52 Charles Ettner
they aim to aid or repair the ills of the body, but not those of the soul. It is the
sociocultural attitudes of a people that make up that soul. And as Sapir once
noted, “[v]ocabulary is a very sensitive index of the culture of a people […]”
(Sapir 1933: 27). Therefore, a reform of the Chinese people’s vocabulary (i.e.,
Chinese language), must also be a part of the struggle. For as long as it is wrong
to say in Chinese ‘women and men are equal’, it is not yet right to
say of Chinese ‘men and women are equal’.
Notes
1. Mainland China officially recognizes eight major dialect groups within the Chinese
language while many of the world’s linguists theorize seven. For a fuller explanation, see
DeFrancis (1984: 37–40), Ramsey (1987: 87–115), and Norman (1988: 181–244).
2. “The term Cantonese, which is sometimes used interchangeably with Yuè, should be
reserved for the dialect of the city of Guǎngzhōu (Canton) and not used as a general name for
the group as a whole.” (Norman 1988: 214)
3. The suffix le and the particle le are presently written with the same Chinese character and
both are pronounced exactly alike, yet they are distinct morphemes, see Ramsey (1987: 55f,
65, 76–78, 272), also DeFrancis (1984: 48) and Norman (1988: 123, 157, 163).
4. Kratochvil (1968: 148) writes: “The traditional Chinese division of characters (that is,
units of an already developed writing system) into six different classes (this division was first
made around 1100 B.C. in the book of classics known as Zhou li ‘The Rites of [the state]
Chou’), does, however, indicate some kind of historical process, since it is based on the
concept that some classes of characters developed from others, their primary classes being
represented by pictures.” See also Norman (1988: 58–61, 67f). The six classes referred to
consisted of 360 pictographs (xiang xing), 80 demonstrations (zhi shi), 110 ideographs (hui
yi), 7,600 ideophones (xing sheng), 80 meta-ideophones (zhuang zhu), and 20,000 loan
characters (jia zhe).
5. Cf. Norman (1988: 60f), Kratochvil (1968: 151f). The phonetic component usually
contributes the medial vowel and final syllabic ending to the pronunciation of the syllable
(not all Chinese words possess final consonants).
6. In ancient Chinese society, the horse mǎ played an essential role in transportation and
in maintaining communication, and it was particularly important to the military forces.
Thus, the ancient Chinese lexicon had a plethora of very specific and detailed terms
concerning horses. Modern Chinese society, however, has much less use for the horse and
thus, much of the lexicon involving horses is now used as verbs such as: qí ‘to ride’, jià
‘to drive’, chí ‘to speed’, along with other word forms pertaining to the movement of
vehicles and human beings (Tang 1988: 59f).
7. The dictionary I used for the survey was Beijing (1988). This dictionary was particularly
appropriate in that it provided for a survey of the more modern Chinese lexicon, given that
Chinese 53
it was published both officially and subsequent to most of the major language reforms in the
People’s Republic of China.
8. One of the more notable dictionaries I surveyed was Mathews (1993).
9. “Pinyin” romanizations appear, with tonemarkers, following the standard four level
scheme for the northern (Mandarin) dialect. Some Chinese graphs have more than one
pronunciation or more than one tone. In either instance, it constitutes a different word with
a different meaning.
10. To describe animals as nán and nu¹ is absolutely unacceptable to Chinese speakers.
Errors of this nature are seldom tolerated and invariably corrected.
11. The writings of Confucius (551–479 BC), a most famous philosopher of ancient China,
discuss at great length notions of propriety of relationships, rights and duties, and many
other moral concepts. These fundamental building blocks of Chinese social order are
contained in a collection of Confucius’ sayings, compiled by his pupils shortly after his death,
now commonly known as ‘The Analects’.
12. Chan (1998: 42) also notes that these factors influence certain areas of Chinese language:
“The Chinese system of kinship terms of reference and address takes such variables as sex,
generation, and lineage into consideration […]”.
13. Optimality theory (cf. Prince & Smolensky 1993) provides a method to illustrate the ordering
of marked and unmarked forms/structures in the context of interacting, ranked constraints.
14. See discussion in Ettner (1993).
15. So unimportant was the Women’s Federation in the government’s eyes that the former
was disbanded in the late 1960s, not to be re-established until 1973 when the main thrust of
the Cultural Revolution had fairly declined.
16. Cf. also Chan (1998: 35): “Studies devoted specifically to issues involving language and
gender were rare. Notable exceptions include Light (1982), probably the very first to
approach the subject and primarily on Cantonese; Shih (1984), the first general overview (in
Chinese); and Farris (1988), the second overview (in English).”
17. The data for these examples was drawn from Tang (1988). On Chinese proverbs see also
Zhang, this volume.
18. Cf. Chan & Madsen & Unger (1984:20, 51, 55–61, 149–151, 157–160, 207–209). Struggle
meetings were brutal, public forums where local individuals were publicly accused of real or
imagined socio-political errors from any given time in their life, and were then required,
under duress, to confess and to castigate themselves. The public’s attendance and participa-
tion was also mandatory.
19. The name wénzì gǎigé wěiyuánhuì is also variously translated as ‘Language Reform
Committee’, ‘Character Reform Committee’, or ‘Writing Reform Committee’.
20. Premier Zhou Enlai outlined the three tasks of the SRC in a January 1958 speech he
delivered to the National Political Consultative Conference “The Immediate Tasks in Writing
Reform”; see Seybolt & Chiang (1979: 5 and 228–243).
21. For a fuller examination of the SRC’s language reforms, see Ettner (1993), Seybolt &
Chiang (1979), Lehmann (1975), and Norman (1988).
54 Charles Ettner
22. Cf. Norman (1988: 81f) and DeFrancis (1984: 257ff).

23. Hodge & Kam (1998: 57) note that “the ‘woman’ radical is a component of many
characters which connote negative values. Throughout this century, language reform
committees have been alerted to this fact, and a group of women have even suggested
changing all such characters by replacing with the ‘evil’ radical . But these pleas have
been consistently ignored on the grounds that while the script may have originally expressed
an ideological stance, it is now beyond ideology. We believe the opposite is the case.”
References
Beijing Foreign Language Institute – English Language Department, ed. 1988. Han Ying Ci
Dian [Chinese-English Dictionary]. Beijing: Southern Commercial Printer’s Book Store.
Chan, Anita & Richard Madsen & Jonathan Unger. 1984. Chen Village: The recent history of
a peasant community in Mao’s China. Berkeley: University of California Press.
Chan, Marjorie K. M. 1998. “Gender differences in the Chinese language: A preliminary
report.” Paper read at The 9th North American Conference on Chinese Linguistics
(NACCL) at Los Angeles.
Cheng, Chin-Chuan. 1986. “Contradictions in Chinese language reform.” International
Journal of the Sociology of Language 59: 87–96.
Croll, Elisabeth J. 1976. The women’s movement in China: A selection of readings. Modern
China Series. Vol. 6. 2nd ed. Nottingham: Russell Press.
DeFrancis, John. 1984. The Chinese language: Fact and fantasy. Honolulu: University of
Hawaii Press.
Ettner, Charles R. 1993. Sexism and the language reforms of the People’s Republic of China:
Socialist language with Chinese characteristics. MA thesis, Stanford University.
Fang, Hanquan & J.H. Heng. 1983. “Social changes and changing address norms in China.”
Language in Society 12: 495–507.
Farris, Catherine S. 1988. “Gender and grammar in Chinese: With implications for language
universals.” Modern China 14: 277–308.
Gao, Yuan. 1987. Born red: A chronicle of the cultural revolution. Stanford: Stanford
University Press.
Hall, Christine. 1997. Daughters of the dragon: Women’s lives in contemporary China.
London: Scarlet Press.
Hodge, Bob & Louie Kam. 1998. The politics of Chinese language and culture: The art of
reading dragons. New York: Routledge.
Honig, Emily & Gail Hershatter. 1988. Personal voices: Chinese women in the 1980’s. Stanford:
Stanford University Press.
Jacka, Tamara. 1997. Women’s work in rural China: Change and continuity in an era of reform.
New York: Cambridge University Press.
Kratochvil, Paul. 1968. The Chinese language today: Features of an emerging standard.
London: Hutchinson University Library.
Lehmann, Winfred P., ed. 1975. Language and linguistics in The People’s Republic of China.
Austin: University of Texas Press.
</TARGET "ett">
Chinese 55
Lien, Chinfa. 1989. “Antonymous quadrinominals in Chinese.” Journal of Chinese Linguis-

tics 17: 263–306.
Light, Timothy. 1982. “On ‘de-ing’.” Computational Analyses of Asian and African Languages
(CAAAL) 19: 21–49.
Mathews, R. H. 1993. Mathews’ Chinese-English dictionary. Revised American edition.
(Originally published as: A Chinese-English dictionary compiled for the China inland
mission by R. H. Mathews, Shanghai: China Inland Mission and Presbyterian Mission
Press, 1931). Cambridge, MA: Harvard University Press.
Norman, Jerry. 1988. Chinese. New York: Cambridge University Press.
Prince, Alan & Paul Smolensky. 1993. Optimality theory: Constraint interaction in generative
grammar. Rutgers University and University of Colorado, Boulder. Unpublished paper.
Ramsey, S. Robert. 1987. The languages of China. Princeton, NJ: Princeton University Press.
Sampson, Geoffrey. 1985. Writing systems: A linguistic introduction. Stanford, CA: Stanford
University Press.
Sapir, Edward. 1933. “Language.” In Selected writings of Edward Sapir in language, culture
and personality, ed. David G. Mandelbaum. 1951. Berkeley, CA: University of Califor-
nia Press, 7–32.
Scotton, Carol Myers & Zhu Wanjin. 1983. “Tongzhi in China: Language change and its
conversational consequences.” Language in Society 12: 477–494.
Seybolt, Peter J. & Gregory Kuei-ke Chiang. 1979. Language reform in China: Documents and
commentary. White Plains, NY: Sharpe.
Shaul, David L. & N. Louanna Furbee. 1998. Language and culture prospect heights. Illinois:
Waveland Press.
Shih, Yu-huei. 1984. “ [A socio-
linguistic study of gender differences in male and female’s language in Chinese].”
Jiaoxue yu Yanjiu 6: 207–229.
Stacey, Judith. 1983. Patriarchy and socialist revolution in China. Berkeley: University of
California Press.
Serruys, Paul. 1962. Survey of the Chinese language reform and the anti-illiteracy movement in
Communist China. Berkeley, CA: Center for Chinese Studies.
Tai, James H.-Y. 1975. “Vocabulary changes in the Chinese language: Some observations on
extent and nature.” Journal of Chinese Linguistics 3: 233–244.
Tan, Dali. 1990. “Sexism in the Chinese language.” Journal of the National Women’s Studies
Association 2: 635–639.
Tang, Ting-Chi. 1988. “ [Male chauvinism in the lexical
structure of Chinese].” In Monographs on modern linguistics: Studies on Chinese
morphology and syntax, ed. Tang, Ting-Chi. Taipei: Student Book Co. 59–65.
Tarantino, Anthony George. 1975. Chinese media of communication: Their history and
political impact in respect to Western developments. Ph.D. dissertation, University of
California, Irvine.
Ting, Pang-hsin. 1969. “ [Tonal relationship
between the two constituents of the dissyllable coordinate construction in Mandarin].”
BIHP XXIX Part II. Taiwan: Institute of History and Philology, Academia Sinica. 155–173.
Wen, Hua. 1973. “On reforming written Chinese.” Peking Review 32: 11–13.
<TARGET "cha" DOCINFO
AUTHOR "Marjorie K. M. Chan"
TITLE "Gender-related use of sentence-final particles in Cantonese"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
chinese
Gender-related use of sentence-final

particles in Cantonese
Marjorie K. M. Chan
The Ohio State University, Columbus, USA.
1. Introduction
2. The Cantonese sentence-final particles je.55, jek.5, ho.35, and wo.33
3. The sentence-final particles la.33, la.55, a.33, and a.55 in two ‘Kalei-
doscope’ episodes
4. Concluding remarks
Notes
References
1. Introduction
In modern colloquial Cantonese, casual conversations among close friends and

relatives abound in sentence-final particles. While some have primarily gram-
matical function, such as turning a declarative sentence into an interrogative
one, most have affective use, reflecting the attitude or emotion of the speaker.
There has been very little research to date on language and gender with respect
to the Chinese language, whether referring to Mandarin, Cantonese, or some
other variety of Chinese (Chan 1996, 1998a, 1998b). This section focuses on the
gender-related use of some sentence-final particles in Cantonese, a variety of
Chinese that is spoken in Guangdong Province in southern China, as well as in
Hong Kong and Macao, and numerous overseas communities all over the
world, including Singapore, Malaysia, Viet Nam, Indonesia, Thailand, the
Philippines, New Zealand, the United Kingdom, the United States, and Canada.
Dialects of the Yue group (which includes Cantonese) are spoken by some 66
million speakers in the world today, based on estimates given in the Summer
Institute of Linguistics’ Ethnologue: Languages of the World (Grimes 1996),
which also ranks Yue as sixteenth in its list of top 100 languages by population.
58 Marjorie K. M. Chan
Standard Cantonese, the prestige subvariety, is spoken in Guangzhou (Canton

City) and Hong Kong.1 It will be shown that general distribution patterns may
obscure some interesting behavior by males with respect to particle usage in
different sociocultural contexts. The research findings emphasize the need to
recognize the complexity of both sociocultural settings and the ramifications of
multiple roles that individuals play in their interaction and communication
with others.
The Cantonese variety of Chinese is particularly interesting to investigate
because it is a tone language with three contrastive level tones in addition to two
rising tones and one or two falling tones (two for those speakers who make a
contrast between a high level tone and a high-falling tone on some lexical
items), and no stress accent system. Hence, every syllable has a tone, including
sentence-final particles. This, it should be noted, is not true of Mandarin
Chinese, where sentence-final particles are toneless, and thus are in the so-
called “neutral tone”. The tonal system in Cantonese, with its contour tones and
tonal register, further curtails the use of intonation to overall raised pitch (Wu
1989: 174ff, cited in Matthews & Yip 1994: 409) and simple rising intonation for
most types of interrogative sentences. As a result, sentence-final particles in
Cantonese often function similarly to intonation in a language such as English,
which has a full-fledged system of stress and no lexical tones. Cantonese also
relies more heavily on sentence-final particles than does Mandarin. The latter
has only four lexical tones, and thus fewer tones to constrain pitch height and
pitch contour, and it has a partially developed stress system. As unstressed
syllables are toneless, they can more readily accommodate pitch changes
demanded by intonation patterns. In contrast to Mandarin, Cantonese has
more sentence-final particles, its speakers use them more frequently in speech,
and produce them with a greater range of pitch and duration differences. Chan
(1998c: 103) notes, for example, that open syllables in Cantonese average 300
milliseconds, while sentence-final particles can have a duration of up to one full
second! When two or three sentence-final particles are concatenated at the end
of an utterance, they can easily exceed one second. This dramatic syllable-
lengthening is not observed in Mandarin. Thus, the striking differences between
these two varieties of Chinese offer great potential for further research with
respect to the study of pragmatic meanings of sentence-final particles in
general, and their gender-differentiated usage in current sociocultural settings
in particular, as part of a larger study of gender across languages.
Sentence-final particles are ubiquitous in colloquial spoken Cantonese.
There are some thirty basic forms (Kwok 1984, Ouyang 1993), which may occur
Chinese (Cantonese) 59
individually or in clusters of two or three at the end of an utterance. Mandarin,

in contrast, has only between seven and at most about seventeen common
sentence-final particles (cf. Matthews & Yip 1994:338, Chao 1968:795ff).
Sentence-final particles serve to add further nuance to what the speaker is saying
beyond the actual content words themselves. Consider, for example, the follow-
ing two pairs of interrogative sentences in Cantonese (extracted from sample
Chinese sentences given in Deng (1991:132)). Transcription here is in Yale
romanization minus tone diacritics. For tones on sentence-final particles, tone
numbers 1 through 5 are used, with 5 for highest pitch and 1 for lowest pitch.
(1) a. Neih sik keuih ma.33?
you know him/her prt
‘Do you know him/her?’
b. Neih sik keuih me.55?
you know him/her prt
‘Do you know him/her?’
(2) a. Neih geisih faan laih a.33?
you when return come prt
‘When are you coming back?’
b. Neih geisih faan laih jek.5?
you when return come prt
‘When are you coming back?’
In the first pair, sentence (1a) is a fairly neutral, information-seeking question.

It is formed by adding the yes-no question particle, ma.33 (mid-level tone), at
the end of the declarative sentence.2 In changing the particle to me.55 (high-
level tone), sentence (1b) conveys the speaker’s startled reaction or surprise.
The context in which it is uttered is not neutral; the speaker is seeking some
kind of confirmation. It might, for instance, be asked by a young woman with
a hint of jealousy to her boyfriend upon seeing an attractive stranger wave to
them. In English, the two sentences are translated identically. One way of
conveying the difference might be the differential use of intonation. Alterna-
tively, using the above scenario, (1b) could be made into an echo question in
English: “You know her?”, and perhaps further adding an interjection: “Oh, you
know her?” Sentence-final particles such as me.55 that have illocutionary force
are not used in formal situations; they are absent in public speeches and news
broadcasts. Affective use of sentence-final particles is generally limited to
informal settings. Moreover, some particles, such as me.55, are found primarily
in casual or intimate conversations among close friends and relatives.
The second pair of sentences consists of two interrogatives. Sentence (2a)

contains the particle a.33 (with phonetic variants, ya.33 and wa.33). It is
optionally added to declaratives, interrogatives, as well as imperatives, function-
ing to soften the tone of voice, thereby conveying greater politeness. This
particle is common to Cantonese and Mandarin in both pronunciation and
pragmatic function. While sentence (2a) is also grammatical without the
particle, its omission makes the sentence sound more abrupt and potentially
more severe. A father might say the sentence without the final particle sternly to
his son as the young man bolts out through the door. Sentence (2b), with jek.5
(high tone) poses a question and at the same time conveys the speaker’s
impatience. Obviously, the two particles, a.33 and jek.5, are not interchange-
able. Use of jek.5 here may also be mixed with some feelings of displeasure, as
the addressee heads out through the door. Sentence (2b) would normally be
used only between people who know each other well, such as between spouses.
And, given traditional gender roles in Chinese culture and society, such a
sentence would stereotypically be viewed as a sentence uttered by the wife
rather than the husband. The pair of sentences in (2) are translated identically
into English, since neither particle has any lexical meaning of its own.
2. The Cantonese sentence-final particles je.55, jek.5, ho.35, and wo.33
Among the thirty or so Cantonese sentence-final particles, je.55 and jek.5

(hereafter simply je and jek) have been studied in detail for gender-related
differences by Chan (1996), based on a survey of twentieth-century textbooks,
dictionaries, and linguistic studies, together with a corpus of twelve videotaped
episodes of a very popular, half-hour weekly television series, Maanfa Tung
‘Kaleidoscope’. Chan (1998a) is a follow-up study of je and jek using the same
corpus. The Kaleidoscope television series was filmed in the mid-1980s by the
Guangdong Television Company on location in Guangzhou (Canton City),
China, in an actual residential area that is referred to in the episodes as Maanfa
Hong ‘Maanfa Lane’. That television series is known for being the first main-
land Chinese television production that used everyday, colloquial Cantonese.
The spirit underlying the series was that it be “as close to real life as possible”.3
As a result, the language used is very natural and full of sentence-final particles.
Chan (1996) suggests that the particles je and jek did not emerge at the
same time. Most likely, the original particle was je, with a delimitative function
and a core meaning of ‘only, merely’. That function is especially obvious in
sentences containing some kind of quantification. The affective use of this

particle by children is noted in Wang (1957: 92), who states that young children
like to use je. It conveys a sense of bragging on the part of the child. Qiao (1966)
may have been the first to identify a gender-related usage of je, noting that it
often occurs as a modal particle in interrogatives in female speech of the kind
that conveys a sense of being affectedly sweet. Decline in usage of this affective
sense, and its gradual replacement by jek is reported by Light (1982),4 who
makes it clear, however, that he does not identify jek as gender-exclusive.
Rather, it is viewed by his informants as a form that is used between lovers,
among family members, or among very close working companions.
While native Cantonese speakers associate jek with female usage, Chan
(1996) observes that both males and females in the Kaleidoscope corpus use je
and jek. Although in absolute terms, males use more je/jek sentence-final
particles than females in the corpus, males use them proportionately less
frequently than females, when one takes into consideration that males produce
a higher proportion of the overall corpus. That is, the twelve episodes contain
a total of 3,657 sentences, of which 1,440 (39.4%) are produced by females, and
2,217 (60.6%) are produced by males. Table 1 (from Chan 1996) gives a total of
206 je and jek sentences in the corpus. As shown, females produce a total of 96
je/jek sentences. This constitutes 6.67% of their total sentence production, or a
1:15 ratio. In the case of the males, even though they produce more je/jek
sentences, their 110 such sentences constitute only 4.96% of their total sentence
production, or a 1:22 ratio. Note also that there is a gender-linked difference in
proportion of je and jek sentences.
Table 1.Distribution of je and jek sentences across gender

Females Males Total
(n = 1440)* (n = 2217) (n = 3657)
JE Sentences 30 55 85

JEK Sentences 66 55 121
Total 96 (6.67%) 110 (4.96%) 206 (5.63%)
*p < .01 level for distribution of JE vs. JEK sentences by females
Females use jek twice as often as they do je, and the results are statistically
significant. Males, on the other hand, produce an equal proportion of both
particles. Furthermore, females’ production of the 66 jek sentences constitutes
4.6% of the total female production of sentences, contrasting with the 55 jek
sentences for males constituting only 2.5% of the total male production of
sentences in the corpus. In other words, females produce jek sentences almost
twice as frequently as males do. The distribution patterns thus clearly support
native speakers’ intuition and perception that jek is more associated with female
speech in today’s spoken Cantonese.
There may also be generational differences in the use of jek. Gao
(1980: 196f), who does not mention je at all, discusses two uses of jek: the first
is to attract the hearers’ attention. The second usage of jek he identifies as
occurring commonly between relatives and very close friends. He further
suggests that jek is used by the younger generation, in speaking to someone of
the older generation, as in children to their parents, and siblings to their elder
brothers and sisters. That usage carries with it some degree of coyness, or
associations of a spoilt child. Gao does not mention gender. Perhaps for girls,
the coyness and affected sweetness of this usage in childhood continues into
early adulthood, and is used with those who have some perceived dominance
over them, in a society that remains patriarchal despite socialist reforms. Boys,
on the other hand, may have been socialized to outgrow that style of speaking,
at least in the public arena. The Kaleidoscope corpus hints at the possibility of
some generation-linked differences in the use of je and jek. Further research is
required to determine whether the differences represent an instance of language
change or stable variation across the generations.
Even though in the series there are no intimate scenes between spouses,
there is an interesting difference between the way men speak to their wives and
how they speak to their friends or neighbors. As shown in Table 2, the wives use
jek proportionately more frequently than je, in fact, almost three times more
frequently than je to their husbands: 37 jek’s to 13 je’s. To their wives, the
husbands also use jek more frequently than je: 18 jek’s to 10 je’s, which is in
sharp contrast to the equal proportion of je and jek in the overall male corpus.
The results suggest that men speak quite differently in private with their
spouses than they do in public settings. A study of the husbands and their
communication with neighbors confirms this. As Table 3 shows, the same
individuals use 27 je’s to 25 jek’s in conversing with their neighbors. The
remaining 30 instances of je and jek (out of a total of 110) are uttered by other
males in the episodes who are either bachelors or not yet of marriageable age.
Correspondingly, for the females, the remaining 39 instances of je and jek (out
of a total of 96) are produced by those who are either widows or not yet of
marriageable age.
What is rarely found in the corpus is the use of jek for affected sweetness or
coyness. Perhaps this is in part due to the absence of truly intimate conversations
Table 2.Distribution of je/jek in wife-husband conversations

Wives to husbands* Husbands to wives
je jek je jek
13 37 10 18
Total 50 28
*p < .01 level for wives’ use of JE vs. JEK to their husbands
Table 3.Distribution of je/jek in wives’/husbands’ conversations with friends

Wives to friends Husbands to friends
je jek je jek
1 6 27 25
Total 7 52
between young lovers or scenes with stereotypical, spoilt daughters and doting
fathers. It may also be part of the egalitarian rhetoric and ideology of post-1949
mainland China, which have discouraged the more strongly feminine-marked
speech and demeanor. Instead, jek is often used to convey a complex mixture of
many different emotions that range from tactfulness and intimate seeking and
sharing of information to those of impatience, exasperation, and dismay.
A follow-up study of the utterances produced by males and females in the
Kaleidoscope corpus (Chan 1998b) reveals that women and men do not
produce proportionately the same amount of declarative sentences and inter-
rogative sentences using je and jek.5 This is shown in Table 4.
Table 4.Distribution of declarative vs. interrogative sentences across gender

Females* Males
Declaratives Interrogatives Declaratives Interrogatives
je 27 3 43 12
jek 5 61 13 42
32 (33%) 64 (67%) 56 (51%) 54 (49%)

Total 96 110
*p < .01
For females, two thirds (67%) of their je/jek productions are in interrogative
sentences and one third (33%) in declarative sentences. Almost all interrogative
sentences (61 out of 64) contain jek rather than je. Males, in contrast, produce
half (49%) of their je/jek utterances as interrogatives, and the other half (51%)
as declaratives. Compared to females, males also show more willingness to utter
jek in declaratives (13 out of 56) and je in interrogatives (12 out of 54). Despite
a pattern of distribution that is not as sharply delineated in male use of je and
jek as in the females’ use of these two particles, the distribution is nonetheless
statistically significant (p < .01).
Some pragmatic functions are neutral, whereas others are linked to males
or to females in the use of je and jek. This is summarized in Table 5 (from
Chan 1998b).
Both women and men use je in declarative sentences for delimiting an amount
or downplaying. They also use it for conveying tact and agreeableness. For gender-
linked usages, females use je in declaratives for coaxing, while males use je in
declaratives where they are refuting something or boasting.
For jek in declaratives, there were no clearly neutral usages. Females use jek
for downplaying, with an added note of exasperation at times. They may also
use jek for saying something tactfully or in a complimentary fashion. Males, on
the other hand, sometimes use jek as a stronger variant of je for downplaying
something, a usage that is absent in females’ production of jek. Males also use
jek in declaratives to express impatience or exasperation (cf. also Fung 2000:50ff).
For interrogatives, je is used almost entirely by males in the corpus, for
reasons that require further research. This has not been noted in the previous
literature: Je is used in information-seeking WH-questions, including those
uttered with impatience, and in tag-questions. On the other hand, both female
and male speakers use jek in interrogatives, and both use it in general informa-
tion-seeking questions as well as in rhetorical questions that contain some
element of sarcasm or exasperation. Besides these gender-neutral uses of jek in
interrogatives, there are also uses that are gender-linked. Females use jek in
more soft-spoken or more intimate information-seeking questions, and in
some rare instances, to express dismay, the most marked female use of jek in
the corpus. For males, in lieu of using jek in interrogatives to show dismay, the
marked usage seems to be that of expressing protest (e.g., a husband’s protest:
‘So many (pieces of stereo equipment), how can I carry all that by myself?!’).
It would normally only be used with someone very familiar or close to the
speaker. This male-linked usage of jek has not been noted in the literature
either. Thus, one finds male-linked uses of je and jek in addition to neutral uses
Table 5.Distribution of gender-neutral and gender-linked uses of je/jek across gender

and sentence types
je jek
1. Declaratives (F: 27, M: 43) (F: 5, M: 13)
Gender- delimiting (‘just, only’) (none)

neutral: downplaying
explaining, clarifying
being tactful, agreeable
Female- coaxing downplaying with exasperation

linked: being tactful, complimentary
intimate relating of news
Male- refuting being impatient, exasperated

linked: turn in course of events jek as a stronger variant of je
boasting for downplaying
2. Interrogatives (F: 3, M: 12) (F: 61, M: 42)
Gender- (none)* general info-seeking**

neutral: rhetorical questions with
sarcasm, exasperation
Female- (none) dismay

linked: intimate info-seeking with
both genders
soft-spoken info-seeking
Male- WH-questions for info- protest and other forms of

linked: seeking complaint
rhetorical questions, incl.
with impatient tone of voice
je embedded in tag-
questions
*There are only two cases of rhetorical questions and one case of je embedded in a tag-question.
**There are three cases of intimate info-seeking by males, but posed to females only.
and those that are female-linked. Moreover, despite the perception that it is
females who are associated with the use of jek, males also use it, albeit to a
much lesser extent, and for different purposes, namely, as a stronger variant of
je and to lodge their protest and unwillingness to perform some task. For
precisely these reasons, neither je nor jek would fit easily into a framework that
would associate the two particles with politeness. It may be plausible to analyze
jek as occurring in contexts of subordination of the speaker, but that usage is
not restricted to females.
In addition to je and jek, another Cantonese particle that has been linked to
gender-differentiated usage is ho.35 (with phonetic variants of smooth or glottal
stop onset), presented in Light (1982: 26), who analyzes ho.35 as a confirmatory
particle that is heard more frequently from women in all stations of life.
However, he notes that it is also heard frequently among men in a lower social
or professional position than their conversational partner. In addition, Light
indicates that ho.35 “is also heard in situations where the speaker is of superior
status and wants to ‘draw out’ his inferior-status listeners by creating an
atmosphere suggestive of free exchange”, and is hence a conscious effort on the
part of the speaker to elicit comments from subordinates.
Light (1982: 30) also discusses the sentence-final particle wo.33, used for
reporting, in which the speaker takes no personal responsibility for the accuracy
of the statement. Here, Light finds that wo.33 is used with noticeable frequency
by domestic servants, secretaries, and clerks in Hong Kong; but these positions
are primarily occupied by women, and hence, “the most that one can say is that
women’s use of wo exceeds that of men in rough proportion to the greater
number of women in social roles which call for utterances using wo”. Light does
identify the pronunciation of this particle with lengthened duration and rising-
falling intonation as gender-linked. That adds a “baby-talk suggestion to the
reporting” and is thus generally perceived as the province of women.
3. The sentence-final particles la.33, la.55, a.33, and a.55

in two ‘Kaleidoscope’ episodes
A systematic study of the Kaleidoscope corpus has not yet been conducted for
particles other than je and jek. Nevertheless, a preliminary study is made of two
of the episodes to determine which additional gender-linked sentence-final
particles might be found in that corpus. Episodes 3 and 8 are chosen because,
for both episodes, there exists not only a romanized transcription of the actual
television production (in Fung 1996a), but also a corresponding Cantonese
script, prepared by the Guangdong Television Company and typed in Chinese
(i.e., monosyllabic, monomorphemic Chinese characters). Of significance is the
fact that the wording in the romanized and character versions of the episodes is
not identical, and the differences are particularly notable with respect to those
innocuous-looking sentence-final particles. In addition to the romanized and
character scripts, the Kaleidoscope project also produced video- and audiotapes
of these two episodes as part of the Kaleidoscope multimedia course materials
(see Note 3).
The study of the sentence-final particles in the two television episodes
yielded some interesting differences in the production by males and females. A
comparison was made between the Cantonese (character) script and the
romanized transcript with respect to what sentence-final particles surface in
Episodes 3 and 8. These two episodes contain a total of 6,088 Chinese characters
(each corresponding to a syllable) in the Cantonese script, and 6,742 syllables in
the romanized transcript. Hence, more syllables are uttered in the television
production than appear in the script, and some of the increase is in sentence-
final particles.
Noteworthy for the comparison of sentence-final particles is the fact that
females in the two episodes speak only a little less than males do: Females
produce 47% of the total corpus in the two Cantonese scripts, a proportion
fairly matched by the 46% in the romanized transcriptions for the actual shows.
Four sentence-final particles are selected for discussion: la.33, la.55, a.33, and
a.55. They occur frequently in the two episodes, and they form a near minimal
set, differing from each other only with respect to presence or absence of a
lateral onset, and high versus mid-level tone.6
The two particles, la.33 and la.55, which differ only in tone, may be paired
together because in the Cantonese scripts, only one Chinese character is used to
represent the two particles, with the performers selecting a mid or high level
tone in the television production. Fung (1996b: 97) identifies two functions for
la.33: The particle may indicate certainty and change of situation (i.e., an
aspectual marker of change of state), or it may be used to tell the listener to do
something, and not simply to suggest it, while she describes la.55 as “a sentence-
final particle implying a casual suggestion or request”.
Table 6 shows that in the two character scripts, females were scripted with
70 instances of the la particle, while males were scripted with 72 of them,
yielding a total of 142 particles, roughly the same number for each gender. On
the other hand, the romanized transcripts, based on transcribing the actual
television production, yield a higher total of occurrences, namely 173, of which
females produced 85 and males 88, again roughly about the same number for
each gender.
From Table 6, a further breakdown is given in Table 7 to show which of the
Table 6.Distribution of la (la.33/la.55) in the scripts across gender

Scripts Females Males Total
Character scripts 70 72 142

Romanized scripts 85 88 173
Total 155 160 315
two particles, la.33 and la.55, are produced by female and male speakers. As a
separation of the two particles cannot be made from the Cantonese script (since
only one Chinese character is used), we focus on the romanized scripts for these
two episodes. For females, of the 85 tokens, about three quarter (74%) repre-
sent the particle la.33 and only one quarter (26%) the particle la.55. Males show
a much less skewed distribution pattern. Although they produce roughly the
same number of tokens, only three fifths (59%) are la.33 and two fifths (36 %)
la.55. Interestingly, the results for females are contrary to stereotypical expecta-
tions that females might produce more of the high-tone particles.
Table 7.Distribution of la.33 and la.55 in the romanized scripts across gender
Romanized scripts Females* Males Total
la.33 63 (74%) 52 (59%) 115

la.55 22 (26%) 36 (41%) 58
Total 85 88 173
*p < .01
In the literature, Kwok (1984: 80f) pairs a.55 with la.55. She notes that in
imperative sentences, there are many cases where these two particles are
interchangeable. In cases involving the first person (singular or plural), a.55
“seems to be more consultative and more lively in tone, inviting the addressee
to agree to the action proposed [… while la.55] does not seem to be so con-
cerned with the addressee’s reactions or feelings”. Again, one may here observe
a potential gender-linked difference that may reflect the stereotypical behavior
expected of women and men. The two Kaleidoscope episodes show a tendency
of gender-linked differences in the production of a.55: Females utter 13 of the
20 tokens and males only 7. With respect to sentence-final particles a.55 and
la.55, the corpus also suggests the possibility of gender-differentiated usage:
a.55 is uttered more frequently by females, and la.55 more so by males. A
larger corpus with more tokens of these two particles is required for a more
detailed investigation.
Finally, the most common sentence-final particle in Cantonese, namely,

a.33, will be discussed and summarized in Table 8. The particle a.33, illustrated
in sentence (2a), serves to soften the utterance, making it less abrupt and less
assertive. In imperative sentences, a.33 lends a more intimate, gentler tone to a
command, as in the case of a parent telling a child to play nicely. If females are
stereotypically expected to be more polite and more soft-spoken, then these
expectations are met in the two episodes. Females produce this particle propor-
tionately more frequently than males: Of the total production of 126 a.33’s,
60% are uttered by females, and only 40% are uttered by males.
Table 8.Distribution of a.33 in the scripts across gender

Scripts Females Males Total
Character scripts 47 (51%) 45 (49%) 92 (100%)

Romanized scripts * 75 (60%) 51 (40%) 126 (100%)
*p < .05
Interestingly, the Cantonese scripts, on the other hand, distributed a.33

almost equally between women and men. Hence, what appears to be a fairly
neutral sentence-final particle based on the Cantonese script reveals gender
differentiation in what is actually said. This sentence-final particle is ubiquitous
and so frequently used by all speakers in colloquial speech that no obvious
gender-linked differences had been noted in the literature.
4. Concluding remarks
In this chapter, several Cantonese sentence-final particles have been presented,

based on what limited research there is to date. The results suggest that females
tend to use sentence-final particles more than males do. Moreover, there are
gender-linked differences in the choice of particles and in their usage in
sentence types. The distribution patterns that emerge in this paper provide a
window into societal expectations of gender-appropriate use of some of the
sentence-final particles in modern Cantonese, and, by extension, provide a
glimpse into gender construction and the socialization process in one region of
China today. The dialogues in the Kaleidoscope series sound natural to viewers
precisely because they conform to stereotypical expectations of male/female
speech behavior.
There remain many issues that need to be addressed in the future; these
include questions of the extent to which gender-linked uses of the particles in
the television series reflect stereotypical uses. Cross-linguistic studies may also
trigger questions on the degree to which cultural differences might play a role
in the amount of gap between stereotypical uses of linguistic forms that might
be found in television series and real-life uses of such forms by the two genders.
It is important to keep in mind that Chinese society has traditionally been
patriarchal, with strongly prescribed gender roles and norms for behavior.
Today, despite the rhetoric of the People’s Republic of China that women “hold
up half the sky”, gender inequality not only continues to exist in educational
opportunities and access to political and economic power, but social, economic
and political changes in China during the 1980’s and 1990’s have increased the
gender gap (cf. Jacka 1997, Chan 1998c). Chinese society, both modern and
traditional, expects conformity, and with such expectations, the gap between
real-life behavior and stereotypical behavior that reflects social norms may turn
out to be narrower than that found in cultures and societies where creativity
and individualism are encouraged.
Notes
1. Differences between the Canton City variety of modern standard Cantonese and that of
the Hong Kong variety are noted in Chan’s (1999) review of Bauer & Benedict (1997), an in-
depth linguistic study of the dialect. For a short historical study of the Yue dialect group to
which Cantonese belongs, see Yue-Hashimoto (1991). For a more comprehensive general
description of Cantonese, cf. Killingley (1993). For a set of prosodic annotation conventions
for Cantonese, see Wong & Chan & Beckman (in press).
2. The most neutral form of the yes-no question is the disjunctive, A-not-A form, e.g., Keuih
sik-mh-sik ngo a.33? (S/he know-not-know me PRT?) ‘Does s/he know me?’. Young Hong
Kong Cantonese speakers, for example, rarely use the ma-particle question form.
3. Thanks go to Professor Xiaobin Jian, who was the first Project Director of the Ohio State
University Cantonese Project (1993–1996), and provided the author with additional
background information on the television series. The author takes this opportunity to thank
her colleague, Professor Galal Walker, Principal Investigator of the OSU Cantonese Project,
for enabling her to use the Kaleidoscope materials for her research. She also thanks Roxana
Fung for discussions and help on the project. Thanks also go to Marlis Hellinger and
Hadumod Bußmann, editors of this set of volumes, Gender across Languages, for their
enthusiasm, dedication, and painstaking efforts.
4. Hence, not surprisingly, studies such as Cheung (1972), Gao (1980), and Matthews & Yip
(1994) only refer to jek and not to je as having affective use. Je is not even mentioned in Gao
(1980), while Cheung (1972) and Matthews & Yip (1994) treat it as primarily a particle with
a delimitative, downplaying function of ‘just, only’. Cheung (1972) allows je/jek as variants
in that context, and includes examples of jek conveying some sense of pride or boastfulness.
5. There were no imperative sentences containing je or jek in the corpus.
6. Neither ho.35 nor wo.33 is included in the two Cantonese scripts although they do surface
in the actual television production of the episodes and in the romanized transcripts.
However, there are only two tokens of ho.35, both uttered by females. The two episodes show
no obvious gender-differentiated usage of wo.33, of which 10 tokens are produced by females
and 12 by males. These preliminary results support Light’s (1982) suggestion that wo.33 is
not gender-linked. There are no instances of wo uttered with rising, or rising-falling, pitch in
the two episodes.
References
Bauer, Robert S. & Paul K. Benedict. 1997. Modern Cantonese phonology. Berlin: de Gruyter.
Chan, Marjorie K. M. 1996. “Gender-marked speech in Cantonese: The case of sentence-final
particles je and jek.” Studies in the Linguistic Sciences 26: 1–38.
Chan, Marjorie K. M. 1998a. “Gender differences in the Chinese language: A preliminary
report.” In Proceedings of the Ninth North American Conference on Chinese Linguistics.
Volume 2, ed. Hua Lin. Los Angeles: GSIL Publications, University of Southern
California, 35–52.
Chan, Marjorie K. M. 1998b. “Sentence particles je and jek in Cantonese and their distribu-
tion across gender and sentence types.” In Engendering communication: Proceedings of
the Fifth Berkeley Women and Language Conference. April 24–26, 1998, Berkeley,
California, eds. Suzanne Wertheim & Ashlee Bailey & Monica Corston-Oliver. Berkeley,
CA: Berkeley Women and Language Group, 117–128.
Chan, Marjorie K. M. 1998c. Review of Matthews, Stephen & Virginia Yip (1994). Cantonese:
A comprehensive grammar. (London: Routledge). Journal of the Chinese Language
Teachers Association 33: 97–106.
Chan, Marjorie K. M. 1999. Review of Bauer, Robert S. & Paul K. Benedict (1997). Modern
Cantonese phonology. (Berlin: de Gruyter). Cahiers de Linguistique Asie Orientale 28:
101–112.
Chao, Yuen Ren. 1968. A grammar of spoken Chinese. Berkeley: University of California Press.
Cheung, Samuel Hung-nin. 1972. Xianggang Yueyu Yufade Yanjiu [Cantonese as spoken in
Hong Kong]. Hong Kong: Chinese University of Hong Kong Press.
Deng, Shaojun. 1991. “Guangzhou Fangyan Changjiande Yuqici [Some common particles of
Cantonese].” Fangyan 2: 126–132.
Fung, Roxana Sukyee, with Teresa Wong & Sandy Wong (compilers). 1996a. Kaleidoscope.
Volume 5. Columbus, Ohio: Foreign Language Publications, Ohio State University.
Fung, Roxana Sukyee. 1996b. A reference guide to spoken Cantonese. Columbus, Ohio:
Foreign Language Publications, Ohio State University.
</TARGET "cha">
Fung, Roxana Sukyee. 2000. Final particles in Standard Cantonese: Semantic extension and
pragmatic inference. Ph.D. dissertation, Ohio State University.
Gao, Huanian. 1980. Guangzhou Fangyan Yanjiu [A study of the Cantonese dialect]. Hong
Kong: Commercial Press.
Grimes, Barbara F., ed. 1996. Ethnologue: Languages of the world. 13th ed. Dallas, Texas:
Summer Institute of Linguistics. (The internet version is online at <http://www.sil.org/
ethnologue/> [accessed 25 January 2000].)
Jacka, Tamara. 1997. Women’s work in rural China: Change and continuity in an era of reform.
Cambridge: Cambridge University Press.
Killingley, Siew-Yue. 1993. Cantonese. München: Lincom Europa.
Kwok, Helen. 1984. Sentence particles in Cantonese. Hong Kong: Centre of Asian Studies.
Light, Timothy. 1982. “On ‘de-ing’.” Computational Analyses of Asian and African Languages
(CAAAL) 19: 21–49.
Matthews, Stephen & Virginia Yip. 1994. Cantonese: A comprehensive grammar. London:
Routledge.
Ouyang, Jueya. 1993. Putonghua Guangzhouhuade Bijiao Yu Xuexi [Comparison and study
of Putonghua and Cantonese]. Beijing: Zhongguo Shehui Kexue Chubanshe.
Qiao, Yannong. 1966. Guangzhou Kouyucide Yanjiu [A study of Cantonese colloquial words].
Hong Kong: Huaqiao Yuwen Chubanshe.
Walker, Galal, series ed. 1994–1997. Kaleidoscope: A course in intermediate to advanced spoken
Cantonese. 5 vols. Columbus, Ohio: Foreign Language Publications, Ohio State University.
Wang, Li. 1957. Guangzhouhua Qianshuo [Elementary introduction to Cantonese]. Beijing:
Wenzi Gaige Chubanshe.
Wong, Peggy W. Y. & Marjorie K. M. Chan & Mary E. Beckman. In press. “An autosegmen-
tal-metrical analysis and prosodic annotation conventions for Cantonese.” In Prosodic
typology and transcription: A unified approach, ed. Sun-Ah Jun. Oxford: Oxford
University Press.
Wu, Kam-yin. 1989. A linguistic study of interrogation in Cantonese: Comparisons with
English. MA thesis, University of Hong Kong.
Yue-Hashimoto, Anne. 1991. “The Yue dialect.” In Languages and Dialects of China [Journal
of Chinese Linguistics Monograph Series 3], ed. William S.-Y. Wang. Berkeley, CA:
University of California Press, 294–324.
<TARGET "zha" DOCINFO
AUTHOR "Hong Zhang"
TITLE "Reality and representation"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
chinese
Reality and representation

Social control and gender relations
in Mandarin Chinese proverbs
Hong Zhang
Colby College, Waterville, Maine, USA
1. Introduction
2. Proverbs as embodiment of traditional gender ideology
3. Challenging the dominant view in proverbial lore
4. Conclusion
Notes
References
1. Introduction
In their concise form and authoritative style, proverbs provide us with rich
linguistic data for the study of the cultural beliefs and social values of a particu-
lar society. In Chinese proverbial lore, almost every aspect of women’s lives is
commented on and evaluated. There are proverbs echoing the dominant social
rules and norms concerning women’s behaviors and familial roles, and proverbs
emphasizing the necessity of male control over women. By contrast, there are
very few proverbs specifically portraying men in stereotypical, male roles or
images.1 Where we have proverbs addressing a male audience, they often appear
as pragmatic advice either telling men which woman to marry, or warning men
to exercise control over women. Women as a group obviously become a marked
category, and are subject to society’s close scrutiny. In this sense, Chinese
proverbs recapture cultural paradigms of gender relations and appear to
reaffirm a gender-related social and cultural hierarchy in a patriarchal and
patrilineal society. However, within the same repertoire of proverbs, we also
find expressions that contradict the dominant cultural ideology of such a
74 Hong Zhang
gender hierarchy, and thus suggest a disjunction between reality and represen-
tation, or the existence of different “realities”. I will argue that contradictions in
representation among various proverbs demonstrate that it is only to a certain
degree that dominant cultural values are shared by the populace at large. And
because they are an oral form of popular culture at the level of local society,
proverbs also provide a means by which gender-based power relations are
renegotiated in everyday life. In the following, examples from Mandarin
Chinese will serve as an illustration of the gendered structures of proverbs (on
Mandarin Chinese, cf. Ettner, this vol.).
2. Proverbs as embodiment of traditional gender ideology
Traditional Chinese society was organized by the patrilineal and patriarchal

kinship system wherein the family name and family estate were transferred
along the male line and marriage was patrilocal.2 Within this male-centered
system, Chinese women inevitably occupied a marginal status. In their natal
home, women were often regarded as temporary members, and a future loss to
their natal family. Upon marriage, they were then viewed as “stranger-intrud-
ers” by their husband’s kin. The low social status of women and negative
attitudes toward them are clearly embodied in the following proverbs:
(1) Jia chuqu de nü, po chuqu de shui.
marry out att daughter throw out att water3
‘A daughter married out is like water thrown out.’
(2) Nü’er shi peiqian huo.
daughter be lose.money commodity
‘A daughter is a commodity on which money is lost.’
(3) Xifu jin le men, quanjia bu an’ning.
daughter.in.law enter pf door whole.family neg peaceful
‘Once a daughter-in-law comes in, there is no peace in the household.’
(4) Xiongdi ru shouzu, qizi ru yifu.
brothers be.like arms.legs wife be.like clothes
‘Brothers are like one’s arms and legs, a wife is like one’s clothes.’
(5) Sheng nan yi dahuan, sheng nü tian youfan.
produce male one big.joy produce female add worry
‘Giving birth to a boy is a great joy, giving birth to a girl adds worries.’
Chinese (Mandarin) 75
(6) Sheng le erzi shi ge bao, shengle nü’er

produce pf son be CL treasure produce daughter
shi gen cao.
be cl weed4
‘To have a son is to obtain a treasure; to have a daughter is to obtain a
weed.’
(7) Erzi shi jinqiang, guniang shi tuqiang.
son be gold.wall daughter be clay.wall
‘Sons are a gold wall, daughters are a clay wall.’
In proverb (1), a woman who is married out is compared to water thrown out,
the implication being that just as one cannot take water back once it is discard-
ed, a woman is of no use to her natal family once she is married. In a straight-
forward way, proverb (2) tells us that one will be on the losing end in raising a
daughter, since eventually she will leave her parents and become useful only to
her husband’s family by perpetuating his patriline. An extended family with
several generations living under one roof was considered the Chinese ideal. But
in reality, joint families were often short-lived, and family division often took
place soon after new conjugal families were formed. Women were often blamed
as the “culprit” who weakened the joint family and caused it to split up. Thus,
in proverb (3), we are told that ‘there is no peace once a daughter-in-law enters
into the household’. The peripheral nature of Chinese women in the Chinese
kinship system is perhaps most graphically described in proverb (4), where a
man’s brothers are said to be his ‘arms and legs’, while his wife is only his
‘clothes’. This proverb warns men that patrilineal solidarity among blood
brothers supersedes the conjugal relation between husband and wife. The
contrastive images in this proverb clearly imply that brothers constitute one
unity in the family and are thus indispensable to each other, just as one cannot
part with one’s arms and legs, whereas a wife assumes only a peripheral place
and can be changed and replaced, just as one can change one’s clothes.
The Chinese patrilineal system inevitably inclined parents to value sons
over daughters, as parents depended on their sons to carry on the family line, to
provide labor for the family, and to support them in their old age. Thus, in
proverb (5), we find that parents are overjoyed when a boy is born but only
grow worried when a girl is born. The contrasting images in proverbs (6) and
(7) are quite telling as well; a son is compared to a ‘treasure’ and a ‘gold wall’,
while a daughter is compared to a ‘weed’ and a ‘clay wall’.
76 Hong Zhang
3. Challenging the dominant view in proverbial lore
The previous proverbs clearly demonstrate that the gender hierarchy in Chinese
society and its kinship system finds its expression in the proverbial lore. In this
way, proverbs seem to validate the dominant cultural values and conform to the
view that women are inferior and subordinate in a patriarchal society. However,
in the same repertoire of proverbs we also find some examples that contradict
the images pictured previously, proverbs that seem to redefine women’s roles in
a positive and constructive way that one rarely sees in other texts and contexts.
One area in which we find women positively depicted is in the role of a wife. In
the Confucian social order of hierarchy, a wife was in a position subordinate to
her husband. The age-old saying fuchang fusui, ‘husband sings, wife echoes’,
could be seen as crystallizing the Confucian ideal regarding the relationship
between husband and wife. However, in the following proverbs, a wife’s
importance is highlighted.
(8) Jia you xianqi, nan’er bu zuo hunshi.
family have good.wife man neg do bad.thing
‘With a good wife at home, the husband does no foolish acts.’
(9) Yang hao yiban gu, qi hao yiban fu.
seedling good one.half crop wife good one.half fortune
‘Good seedlings guarantee half of one’s crops; a good wife guarantees
half of one’s fortune.’
(10) Miao zhong bu hao yinian qiong, qu qi
seedling plant neg well one.year poor get wife
bu hao yishi qiong.
neg well one.life poor
‘A man is poor only for one year if he does not plant his seedlings well;
he is poor all his life if he does not marry a good wife.’
(11) Long li wu ji bu zhi tian liang,
chicken.house inside neg rooster neg know sky bright
chuang shang wu qi mei ren shangliang.
bed on neg wife neg person consult
‘If there is no rooster in the chicken house, one does not know when the
dawn breaks; if there is no wife in bed, one has nobody to consult with.’
In proverb (8), a good wife is said to make a man become a good person, while
proverb (9) emphasizes that just as good crops depend on good seedlings, a
man’s fortune depends on him having a good wife. This same message is
reflected in proverb (10), where the parallel comparison suggests that a man
only suffers a one-year loss if he fails to plant his crops well, but that he will pass
his whole life in poverty if he fails to marry a good wife. There are clearly two
readings of the above proverbs. On the one hand, these proverbs recognize the
vital role a woman can play in a man’s life and give credit to such a role. On the
other hand, as is inevitable in a patriarchal society, the importance and social
value of a woman are defined and acknowledged only through her usefulness
and service to her husband. However, if proverbs (8), (9), and (10) still define
the importance of a wife from a man’s point of view, proverb (11) vividly
conveys the message that a man’s wife is his significant other; here the wife is
treated as an equal partner, a confidant whose opinion is sought after. More-
over, this proverb also reveals to some extent male vulnerability and depen-
dence which are usually not acknowledged. In a male-dominant society like
traditional China, such a proverb truly voices a different perspective on the
interdependent nature of husband and wife in their daily life.
Another area of contradiction between reality and representation in
proverbs has to do with the conjugal relationship. In the Chinese patrilineal
system, the conjugal relationship between husband and wife is downplayed,
while the vertical parent-child relationship is emphasized. As Levy (1949: 110)
points out, “throughout Chinese society, the consideration of feelings went in
one direction – from children to parents, from youth to their elders. One did
not concern oneself with the son’s happiness; his father’s happiness was what
mattered”. Filial piety not only constituted a fundamental ethical concept of
Confucianism, it had manifestations in all aspects of Chinese life from the
process of socialization to the practice of ancestor worship to the regulation of
state laws. The classical collection of The Twenty-Four Examples of Filial Piety
records a filial story about a man who expelled his wife on the grounds that she
did not pay proper worship to his ancestors (Tang 1844).5 According to Francis
Hsu (1948), in Chinese society, “the husband-wife relationship is strictly held
to be supplementary and subordinate to the parent-son relationship”, and the
development of a close relationship between woman and man is strongly
discouraged because it is considered “detrimental to the supremacy of filial
piety” (Hsu 1948: 57, 209). The distance between husband and wife has to be
maintained for the sake of the vertical parent-son relation and patrilineal unity.
From a psychological perspective, Margery Wolf (1972) argues that conjugal
affection was discouraged in a Chinese family because it endangered the mother-
in-law’s sense of security and threatened to dismember her “uterine family”.
However, peasant proverbs undermining such dominant ideologies
abound; here the conjugal relation is exalted and celebrated. Proverb (12)
78 Hong Zhang
emphasizes the bonding effect of husband and wife, while proverb (13) high-
lights the importance of unity between husband and wife. In proverb (14),
conjugal love is even placed ahead of filial piety.
(12) Yi ri fuqi bai ri en,
one day husband.wife hundred day devotion
bai ri fuqi si hai shen.
hundred day husband.wife as sea deep
‘One day together as husband and wife brings a hundred days of devotion
to each other, a hundred days together brings devotion as deep as the sea.’
(13) Jiaren bu he linshe qi,
family.member neg harmony neighbor bully
fuqi bu he liuchu qi.
husband.wife neg harmony domestic.animal bully
‘When family members do not unite, they will be bullied by their neighbors;
when couples do not unite, they will be bullied by their domestic animals.’
(14) Fumu en’ai zhong you bie,
parents gratitude.love eventually have separation
fuqi yishi ai mo fen.
husband.wife whole.life love neg parting
‘No matter how deep one’s gratitude toward his/her parents, it is going to
end; but the tie between husband and wife is as everlasting as a lifetime.’
In traditional China, having many children (especially sons) was considered

good fortune, since, among other things, a person’s old age support was guaran-
teed; as the saying goes yang er fanglao ‘to raise sons is to secure one’s old age.’
On the other hand, a second marriage was regretted as a misfortune, as it
indicated that a person’s first marriage failed to last his/her entire life. However,
by nonetheless favoring a second marriage, proverb (15) foregrounds the
conjugal bond and the interdependence between husband and wife.
(15) Man tang ernu bi bushang banlu
whole house son.daughter comp neg half.way
fuqi.
husband.wife
‘Even a second marriage is better than a houseful of children.’
4. Conclusion
This study has shown that as an oral form of popular culture, Chinese proverbs
on the one hand reaffirm the gender hierarchy legitimized by the dominant
male-centered cultural system, and thus appear to lend support to the view that
folklore validates culture (Bascom 1965). However, on the other hand, there are
also proverbs that are in opposition to the norms and values advocated by the
orthodox elite, indicating that at the level of local society, the hegemonial
control of the state culture was only partial and that peasant acceptance of
orthodox values was limited. According to Elmslie (1917), contradicting
proverbs reflect different worldviews by different classes of people, indicating
the heterodox nature of proverbs. In his study of twentieth-century Chinese
peasant proverbs, Arkush (1990) finds that both elements of orthodoxy and
heterodoxy exist; while some proverbs reflect shared ideas and beliefs that are
in accord with the elite values of Confucian orthodoxy, some other proverbs
suggest “a spirit of resentment toward the elite, a sort of proto-class conscious-
ness or implicit questioning of the legitimacy of the orthodox ideas of social
hierarchy” (Arkush 1990: 329).
The contradictions in the proverbs discussed previously suggest that
although Chinese women occupied an inferior social position within the
dominant cultural hierarchy, in everyday life experiences gender relations might
have been much more fluid and variable than what is implied by the male-
centered social system. The images of women as worthy and competent
partners, and of a more intimate relation between spouses might have deviated
from the norms of cultural expectation, they nevertheless constituted the living
social reality to a certain degree. Owing to their oral nature utilized mostly by
illiterate peasants, proverbs thus allow us to view some of the concrete cultural
practices which demonstrate that peasant culture partakes of the elite culture,
sharing its norms and values. At the same time, these practices also constitute
an arena that provide room for local variations, renegotiations, and challenges
to elite norms.
Notes
1. I have surveyed three large collections of Chinese proverbs: Suyu [Popular Sayings]
(Beijing Folklore Editorial Board 1983), Zhonghua Yanyu Zhi [Records of Chinese Proverbs]
(Zhu 1989), and Zhongxiang Minjian Wenxue Zuopin Xuanji: Yanyu Fence [Collection of
Zhongxiang County Folklore: Proverbs] (Zhongxiang Folklore Board 1991). These three
</TARGET "zha">
80 Hong Zhang
volumes contain 40,000 to 60,000 Chinese proverbs, of which approximately 500 proverbs
deal with women and their familial roles and social behaviors specifically, while there are less
than 30 proverbs specifically depicting men and the male gender role.
2. During the second half of the 20th century, great social and economic changes have taken
place in China. Although today in rural China most marriages are still patrilocal, the
patrilineal kinship system which is the basis of traditional Chinese society has almost
disappeared. With more and more people now living in nuclear families, women’s status at
home and in the community has increased considerably. Throughout this chapter, the past
tense is used to indicate that this is how the traditional social order was reflected and re-
enacted in proverbs dealing with gender relations.
3. The element de should be characterized as att (attributive) rather than gen (genitive).
4. The reason why there are so few classifiers in the proverbs listed here is that proverbial
language is more succinct and elliptical. While in ordinary language many nouns will carry
a classifier, in a proverb the same nouns may not function as countable or individualized, but
become uncountable or abstract, and therefore do not carry a classifier.
5. The Twenty-Four Examples of Filial Piety (known as Er-shi-si xiao) is a collection of stories
citing examples of exceptional sacrifices made by children for their parents. The earliest
collection was circulated during the Song Dynasty (980–1280), and since then it has been
revised and edited constantly as one of the most widely used sources for teaching children
filial piety and proper behavior.
References
Arkush, R. David. 1990. “Orthodoxy and heterodoxy in twentieth-century Chinese peasant

proverbs.” In Orthodoxy in late imperial China, ed. Kwang Ching Liu. Berkeley, CA:
University of California Press, 311–335.
Bascom, William. 1965. “Four functions of folklore.” In The study of folklore, ed. Alan
Dundes. Englewood Cliffs, NJ: Prentice Hall, 279–298.
Beijing Folklore Editorial Board. 1983. Suyu [Popular sayings]. Beijing: Beijing Folklore Press.
Elmslie, W. A. L. 1917. Studies in life from Jewish proverbs. London: Charke.
Hsu, Francis L. K. 1948. (Rev. and enlarged 1971) Under the ancestors’ shadow: Kinship,
personality, and social mobility in village China. Stanford, CA.: Stanford University Press.
Levy, Marion Joseph, J. R. 1949. (reprint 1963). The family revolution in modern China. New
York: Octagon.
Tang, Bi, ed. 1844. (reprint 1991). Qianhou xiaoxing lu [Two volumes of filial acts (also
known as) The twenty-four examples of filial piety]. Shanghai: Shanghai Wenyi Press.
Wolf, Margery. 1972. Women and the family in rural Taiwan. Stanford, CA: Stanford
University Press.
Zhongxiang Folklore Board. 1991. Zhongxiang Minjian Wenxue Zuopin Yuanji: Yanyu Fence
[Collection of Zhongxiang county folklore: Proverbs]. Beijing: Chinese Folklore Press.
Zhu, Jiefan. 1989. Zhonghua Yanyu Zhi [Records of Chinese proverbs]. Taipei: Taiwan
Commercial Press.
<TARGET "ger" DOCINFO
AUTHOR "Marinel Gerritsen"
TITLE "Towards a more gender-fair usage in Netherlands Dutch"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
dutch
Towards a more gender-fair usage

in Netherlands Dutch
Marinel Gerritsen
University of Nijmegen, The Netherlands
1. Introduction
2. Selected structural properties of Dutch
2.2 Gender-specific vs. gender-indefinite personal nouns
2.3 The morphology of personal nouns
2.3.1 Derivation
2.3.2 Compounding
3. Gender-biased usage: Variation and tendencies of change
3.1 Historical background
3.2 Generic masculine nouns and pronominalisation
3.2.1 Problems in traditional usage
3.2.2 Tendencies of variation and change
3.3 Terms of address and family names
3.3.1 Terms of address
3.3.2 Family names
3.4 Idiomatic expressions and stereotypical descriptions of women
and men
3.5 Occupational terms: A case of Dutch language politics
3.5.2 The recommendations and the public debate
3.5.3 The manual for professional terms of the Dutch Language
Union (2000)
4. The use of occupational titles in advertising: An empirical study
5. Conclusion
Notes
References
82 Marinel Gerritsen
1. Introduction
Modern Standard Dutch (Nederlands) is the official language of the Nether-

lands and one of the official languages of Belgium (cf. Kooij 1987). In the
Netherlands, the number of speakers of Dutch is approximately 15.4 million,
while ca. 4.6 million use it in Belgium. Dutch is also the official language of
Suriname and the Dutch Antilles. In addition, Dutch is spoken by smaller
groups of speakers in Australia, Indonesia and elsewhere (cf. de Vries &
Willemyns & Burger 1993). In the Republic of South Africa, a descendant from
17th century Dutch has developed into Afrikaans, which is now regarded as a
separate language.
Dutch belongs to the West Germanic branch of the Germanic languages.
Early written documents date from the period of Middle Dutch (1100–1500).
Most of these were written in the dialects of the leading southern provinces,
Flanders and Brabant. Modern Dutch developed from the dialects spoken in the
western part of the Netherlands and the Brabantian area in Belgium. Compared
to German (and similar to English), Modern Dutch has lost most of its case
distinctions and inflectional morphology.1
Gender bias in the varieties of Dutch as used in Suriname, the Dutch
Antilles and Indonesia is politically and linguistically so different from the
situation in the Netherlands that it cannot be dealt with here. Nor can the issue
of gender bias in Dutch as it is used in Belgium be discussed. In Belgium,
Dutch competes with French and German and is often called Flemish, or
Vlaams in Dutch. There are some differences in morphological structure
between Belgian and Netherlands Dutch that especially affect gender, so that
the Belgians have partly used other linguistic and political means to solve the
problem of gender bias.2
2. Selected structural properties of Dutch

Dutch holds an intermediate position between English and German regarding
grammatical gender distinctions in nouns: It has more distinctions than
English, but fewer than German (Kooij 1987: 145).3 Nouns can be divided into
two classes: nouns which in the singular take the definite article de, and nouns
which take the definite article het. Nouns belonging to the de-class are either
Dutch 83
masculine or feminine. For instance de tafel ‘the table’ is feminine and requires
the feminine personal pronouns zij ‘she’ and haar ‘her’, and the feminine
possessive pronoun haar ‘her’ (cf. 1a). De stoel ‘the chair’ is masculine and
requires the masculine personal pronouns hij ‘he’ and hem ‘him’, and the
possessive pronoun zijn ‘his’ (cf. 1b). Nouns in the het-class are neuter, e.g., het
bed ‘the bed’; they require the same possessive pronoun as masculine nouns.
They take two personal pronouns, either the masculine personal pronouns or
het ‘it’ (cf. 1c). For nouns in the plural no gender distinctions are made: The
article for all three genders is de, the possessive pronoun is hun ‘their’, and the
personal pronouns are zij ‘they’ and hen/hun/ze ‘them’ (cf. 1d). The factors
conditioning the variation between hen/ze and hun are complex, but not
relevant to gender.
(1) a. De tafel (f) met haar poten. Zij is mooi. Ik zie haar.
‘The table with its (her) legs. It (she) is beautiful. I see it (her).’
b. De stoel (m) met zijn poten. Hij is mooi. Ik zie hem.
‘The chair with its (his) legs. It (he) is beautiful. I see it (him).’
c. Het bed (n) met zijn poten. Hij/het is mooi. Ik zie hem/het.
‘The bed with its (his) legs. It (he/it) is beautiful. I see it (him/it).’
d. De tafels met hun poten. Zij zijn mooi. Ik zie hen/ze/hun.
‘The tables with their legs. They are beautiful. I see them.’
Table 1 summarises the use of articles, possessive and personal pronouns for
masculine, feminine and neuter nouns in both singular and plural. It is clear
that the distinction between the genders is only represented in the singular. The
difference between masculine and feminine nouns is expressed in possessive
and personal pronouns (but see also Section 3.2), and the difference between
neuter and masculine nouns is expressed in the definite article.
The distinction between neuter gender on the one hand, and masculine and
feminine gender on the other, is expressed in the singular in two other word
classes: demonstratives and adjectives. Dat ‘that’ and dit ‘this’ (dat bed ‘that
bed’, dit bed ‘this bed’) are neuter gender, while die ‘that’ and deze ‘this’ (die
stoel ‘that chair’, die tafel ‘that table’, deze stoel ‘this chair’, deze tafel ‘this table’)
are masculine and feminine. The attributive adjective ends in -e before nouns
of all genders, singular and plural. The only exception occurs with singular
neuter nouns in indefinite contexts, where the adjective is uninflected (een
nieuw bed ‘a new bed’).
Grammatical gender is not expressed by other means in Dutch, neither in
verb agreement as in a number of Romance languages, nor in case inflection as
Table 1.Grammatical gender distinctions in Dutch

Def. article Poss. pronoun Pers. pronoun
Nom Acc/Dat
Singular
Masculine de zijn hij hem
Feminine de haar zij haar
Neuter het zijn hij hem
Plural
Masculine
Feminine de hun zij hen/ze/hun
Neuter
in German. Nominative, dative and accusative case are not expressed in Dutch
nouns. The distinction is only expressed in personal pronouns (cf. Table 1) and
in some archaic expressions. Again, the Dutch situation is in between that of
English and German. In the English language, neither case nor gender is
expressed, while both case and gender are expressed in German (cf. 2a, b). In
Dutch, the distinction between neuter gender and masculine/feminine gender
is expressed in the article, but case is not:
(2) a. The woman gave the apple to the child.
b. Die Frau gab dem Kind den Apfel.
c. De vrouw gaf de appel aan het kind.
2.2 Gender-specific vs. gender-indefinite personal nouns

Dutch has nouns with lexical gender, such as vrouw ‘woman’, echtgenote ‘wife’,
mevrouw ‘Mrs’, juffrouw ‘Miss’, wijf ‘bitch’, tante ‘aunt’ or nicht ‘niece, cousin’ for
females (which are feminine) and man ‘man’, echtgenoot ‘husband’, meneer ‘sir’,
oom ‘uncle’ or neef ‘nephew, cousin’ for males (which are masculine). However,
grammatical gender does not always coincide with lexical gender. Thus, meisje
‘girl’ and jongetje ‘boy’ are female and male, respectively, but grammatically neuter;
of course, both are diminutives carrying the respective derivational suffix.
Not all professional titles in Dutch have a masculine and a feminine
counterpart, which is different, e.g., from German. Dutch has two types of
professional titles: terms that indicate the gender of the person who practises
the profession and terms that do not.
Dutch 85
We can distinguish two categories of professional terms that have a male

and a female counterpart. First, there are terms that semantically differ only in
the referential gender that is indicated:
(3) Terms for women Terms for men Translation
[Feminine] [Masculine]
lerares leraar ‘teacher’
schrijfster schrijver ‘author’
actrice acteur ‘actress/actor’
componiste componist ‘composer’
assistente assistent ‘assistant’
Second, there are terms that not only indicate a difference in referential gender,
but also include other semantic asymmetries:
(or men/women)
secretaresse secretaris ‘secretary’
masseuse masseur ‘masseur’
directrice directeur ‘director’
caissière kassier ‘cashier’
In most cases the professional term indicating a woman refers to a job with a
lower social status than the term indicating a man. A secretaresse earns far less than
a secretaris. A secretaris very often has a secretaresse, who works for him/her but a
secretaresse never has a secretaris who works for her. There is a clear difference in
function, and women who have the function of secretaris will never call themselves
secretaresse. Examples of men who have the function of secretaresse hardly occur.
As far as I know men with this function call themselves administratief mede-
werker ‘white-collar worker’. Similarly, the term masseuse is associated with a
whorehouse, but the term masseur with a physiotherapeutic centre. A directrice
can be the head of a pre-school kindergarden, or home for the elderly, but when
a woman becomes the director of a grammar school or a large organisation she
calls herself directeur. The caissière works in a store, the kassier in a bank.
Furthermore, Dutch has professional terms that either do not have any
equivalents referring to men, as in (5a), or equivalents referring to women
(5b), although the morphological equivalents could be built. In both cases,
new names are invented when men enter the jobs mentioned in (5a), for
example huishoudelijke hulp ‘domestic help’ for werker, and when women enter
the jobs in (5b).
(5) a. Terms for women Terms for men Translation

kleuterleidster *kleuterleider ‘nursery school teacher’
vroedvrouw *vroedman ‘midwife’
winkelmeisje *winkeljongen ‘sales girl’
naaister *naaier ‘dressmaker’
werkster *werker ‘cleaning lady’
b. *melkboerin melkboer ‘milkman’
*timmervrouw timmerman ‘carpenter’
A fourth category are the so-called “gender-neutral” terms which are grammati-
cally masculine:
(6) Terms for women and men Translation
dokter ‘doctor’
professor ‘professor’
psychiater ‘psychiatrist’
consul ‘consul’
bediende ‘servant’
beambte ‘official’
notaris ‘notary’
minister ‘minister’
informaticus ‘information scientist’
fysicus ‘physicist’
ingenieur ‘engineer’
dominee ‘clergyman’
These are terms to which a feminine suffix cannot easily be added. This is partly
due to the fact that the addition of such suffixes would result in homonyms of
existing (usually non-personal) terms, e.g., informatica ‘information science’,
fysica ‘physics’, or in forms which are already in use for the wives of the men
with the said profession, as in domineese ‘clergyman’s wife’.4 In addition, it
seems that female-specific suffixes cannot easily be added to these terms
because they are loanwords.
2.3 The morphology of personal nouns

2.3.1 Derivation
As indicated in Section 2.2, not all personal masculines have a feminine
counterpart. In this respect Dutch differs substantially from German, where
Dutch 87
almost all personal nouns referring to men can be transformed into feminine
equivalents by adding the suffix -in (cf. Bußmann & Hellinger, vol. III). In
Dutch, the number of suffixes that can be used to feminize personal nouns
denoting men is much larger than in German. The following list of female-
specific suffixes starts with the productive suffixes and ends with the unproduc-
tive suffixes. The data presented here are based on ANS (1997: 668–672) and
Brouwer (1985).
The suffix -e. Feminine personal nouns can be derived from masculine
personal nouns by adding the suffix -e. This productive process especially
occurs with loans (7a), nouns ending in -ing (7b) and some other nouns (7c):
(7) Masculine terms Feminine terms Translation
a. assistent assistente ‘assistant’
spion spionne ‘spy’
student studente ‘student’
advocaat advocate ‘lawyer’
b. leerling leerlinge ‘pupil’
c. echtgenoot echtgenote ‘spouse’
gast gaste ‘guest’
The suffix -ster. Masculine personal nouns that are derived from verbs can be
transformed into feminine personal nouns by changing the suffix -er to -ster (cf.
8a). The suffix -ster is also used to transform masculine personal nouns ending
in -aar or -ier into feminine nouns (cf. 8b). Both derivational processes are
productive.
a. arbeider arbeidster ‘worker’
voorzitter voorzitster ‘chairperson’
b. bedelaar bedelaarster ‘beggar’
tuinier tuinierster ‘gardener’
The suffixes -euse and -trice. Masculine personal nouns ending in -eur or -tor
can be transformed into feminine nouns by replacing -eur with -euse (9a) and
-tor with -trice (9b). This process is productive; however, both masculine
suffixes only occur in loanwords which are not frequently borrowed today.
a. adviseur adviseuse ‘advisor’
chauffeur chauffeuse ‘chauffeur’
b. illustrator illustratrice ‘illustrator’
The suffix -a. Masculine personal nouns ending in -us can be transformed into
feminine nouns by changing -us to -a. This process is still productive, but the
suffix -us only occurs in loanwords from Latin, and today not many Latin words
enter the Dutch language.
(10) Masculine term Feminine term Translation
historicus historica ‘historian’
The suffixes -es and -esse. Feminine personal nouns may be derived from
masculine nouns by adding the suffixes -es or -esse (11a) or by changing the
suffix -is of masculine personal nouns into -esse (11b). Both processes are no
longer productive in Dutch. For some feminine personal nouns, two forms
exist. In those cases the -esse-variant is more prestigious than the -es-form.
a. baron barones, baronesse ‘baron, baroness’
prins prinses ‘prince, princess’
b. archivaris archivaresse ‘archivist’
jubilaris jubilaresse ‘person celebrating a jubilee’
The suffix -in. Derivation by -in from masculine personal nouns is no longer
productive.
boer boerin ‘farmer’
held heldin ‘hero’
The suffix -se. Derivation by -se is no longer productive. Respective feminine

personal nouns are archaic. Most of these words not only denote a female
person, but have the additional meaning ‘wife of ’, except for kasteleinse, which
has both meanings: ‘female innkeeper’ and ‘wife of innkeeper’.
dominee domineese ‘wife of minister’
schipper schipperse ‘wife of shipmaster’
kastelein kasteleinse ‘(wife of) innkeeper’
Since the Middle Ages, substantial changes have taken place in the derivation of
feminine personal nouns from masculine personal nouns (cf. Brouwer 1980,
1985). Both the number of suffixes that can be used to this end – Middle Dutch
had three other suffixes (-egge, -sche, -nede) – and the degree of productivity
have decreased: The currently unproductive suffixes -in and -es/-esse were
Dutch 89
productive in Middle Dutch, and -ster was more productive than it is in current
Dutch (cf. also the parallel development in Norwegian described by Bull &
Swan, this vol.). The possibilities of deriving feminine nouns from masculine
nouns in Dutch have generally decreased.
2.3.2 Compounding
Compounding is another process by which masculine personal nouns can be
feminised. A noun that has the word man ‘man’ as a last part can be trans-
formed into a noun denoting a woman by replacing man with vrouw ‘woman’
(cf. 14). However, it was indicated in Section 2.2 that not all masculine personal
nouns ending in -man can be changed into a feminine personal noun.
bloemenman bloemenvrouw ‘flower-seller’
cameraman cameravrouw ‘cameraman, camerawoman’
Words such as persoon (m) ‘person’ are rarely used in the singular as a neutral
alternative. However, in the plural personen or mensen ‘people’ are sometimes
used as such: ambtspersonen ‘officials’, zeemensen ‘sailors’ (lit. sea-people).
3. Gender-biased usage: Variation and tendencies of change
3.1 Historical background

The fight for a less gender-biased Dutch language dates from the beginning of
the second feminist movement, which in the Netherlands is marked by the
appearance of Kool-Smit (1967), an article entitled Het onbehagen bij de vrouw
‘The discontentment of women’, published in the prestigious Dutch literary
journal De Gids. Kool-Smit (1933–1981) showed that women were considered
as secondary to men in nearly all sectors of life. Her passionate plea for equality
centred around three main points:
a. The fair division of “inside” and “outside” services (household and
professional life);
b. Redistribution of power between women and men;
c. The removal of the segmentation between women’s and men’s professions.
The impact of this article on Dutch women was enormous. Many appeared to
have the same feelings and joined forces in several feminist unions. The two
most important women’s liberation movements were the Dolle Mina’s ‘Crazy
Minas’ and Man Vrouw Maatschappij ‘Men Women Society’, which were
founded by Kool-Smit. Whereas the former fought against inequality by means
of happenings such as tying purple ribbons to pissoirs, Man Vrouw Maat-
schappij used more serious means: Their members exposed the inequality by
analysing socio-economic policies and legislation of the Dutch government. It
was also this movement that made suggestions for gender-fair professional
titles in Dutch.
However, during the first years of the second feminist movement gender-
biased language was not an issue in the Netherlands. It was only in 1975 that
Annie Romein-Verschoor (1895–1978), a historian and specialist in Dutch,
published a comprehensive article about the second feminist movement in
which she also paid attention to linguistic problems. She mentioned three
different fields: gender bias in the usage of Dutch, reference to generic nouns
with masculine pronouns and gender-specific terms for professions.
Most aspects of gender bias in Dutch to which Romein-Verschoor (1975)
drew attention were discussed vehemently in the Netherlands in subsequent
years. It is striking that she herself does not make any suggestions for change.
She demonstrates that gender bias in Dutch was largely a symptom of the
inequality in Dutch society, but she also emphasised that this society is chang-
ing and that the Dutch language lags behind the social developments. She warns
that the gender bias in Dutch might slow down the progress of women’s
liberation and states that it is therefore important to find solutions, which, of
course, is not easy to effectuate: “It is easy to protest, but to find solutions is
difficult because language develops in its own way and is not determined by
committees of linguists” (Romein-Verschoor 1977: 14).
Although Romein-Verschoor did not herself solve the problem of gender
bias in Dutch, she was the prime mover. It was, however, not her comprehen-
sive 1975 article in De Gids, which also dealt with many other problems of
sexism, but a paper she had planned to deliver at the 1977 meeting of the
Vereniging voor Vrouwen met een Wetenschappelijke Opleiding (VVAO) ‘Society
for Women with an Academic Education’ (Romein-Verschoor 1977). Regard-
ing gender bias in language the content of this paper was similar to that of
Romein-Verschoor (1975). The executive committee of the VVAO considered
the topics raised in this presentation so important that they placed a request in
the feminist magazine Opzij in which they asked linguists to form a committee
to find solutions to the gender bias in Dutch as observed by Romein-Verschoor.
A number of women – among others Agnes Verbiest, Dédé Brouwer and the
Dutch 91
author of the present article – answered this call and set to work. Apart from
this group, which was called Research Group Language Sexism of the VVAO,
other people, both linguists and non-linguists, also dealt with one or more of
the three different aspects of gender bias in Dutch as raised by Romein-Ver-
schoor. At the time the topic of “language and sex”, as it was called, was popular
among students of linguistics. This was partly due to Lakoff (1975) and devel-
opments in sociolinguistics, which increasingly demonstrated that women and
men differed in language use. Although this was a different topic than gender
bias in language, it raised the interest in sexism in language.
Below the major issues in the debate will be discussed: generic masculine
nouns and pronominalisation, terms of address and family names, idiomatic
expressions and stereotypical descriptions of women and men, and occupa-
tional terms.
3.2 Generic masculine nouns and pronominalisation

For nouns denoting animate entities, grammatical and referential gender
sometimes coincide: vrouw ‘woman’ and weduwe ‘widow’ are feminine, while
man ‘man’, weduwnaar ‘widower’ are masculine. The choice of anaphoric
pronouns presents no problems in these cases:
(15) a. de weduwe en haar kinderen
‘the widow and her children’
b. de weduwnaar en zijn kinderen
‘the widower and his children’
However, for a large number of nouns, referential and grammatical gender do

not correspond. There are gender-indefinite nouns that refer to both female
and male beings: persoon (m) ‘person’, kind (n) ‘child’, arts (m) ‘doctor’, where
the grammar of Dutch requires masculine personal/possessive pronouns when
they are used generically, with the consequence that referential gender remains
ambiguous. This also applies to indefinite pronouns such as iedereen ‘every-
body’. Due to the grammatical structure of Dutch and prescriptive attitudes
(ANS 1997: 230–235, Woordenlijst Nederlandse Taal 1995: 41–43), women are
not visible in these cases:
(16) a. de mens (m) en zijn kinderen
‘man and his children’
b. Een arts (m) moet naar zijn patienten luisteren.

‘A doctor must listen to his patients.’
c. Iedereen (m) wil graag begrijpen wat hij leest.
‘Everybody wants to understand what he reads.’
3.2.2 Tendencies of variation and change

No solution has been found for the problem of masculine pronominal agree-
ment with potentially generic antecedents in contexts in which the referents’
gender is either not known or unimportant. Some have proposed using the
word zaar, a mixture of feminine haar and masculine zijn in sentences such as
(17). However, this alternative is not frequently used.
(17) de mens en zaar kinderen
‘the human being and his/her children’
Instead of trying to introduce new gender-neutral pronouns, other solutions

have been proposed (cf. Renkema 1995 and Van Gessel et al. 1992): the use of
plurals, since plural pronouns have no gender distinction (cf. 18), the use of
pronominal splitting (19), or the use of a paraphrase (20):
(18) de mensen en hun kinderen
‘humankind and their children’
(19) Iedereen wil graag begrijpen wat hij of zij leest.
‘Everybody wants to understand what he or she reads.’
(20) Wie leest, wil ook begrijpen.
‘Who reads wants to understand.’
However, if people feel that they cannot avoid masculine pronouns, they
sometimes mention in a note that hij ‘he’, hem ‘him’ and zijn ‘his’ are supposed
to include zij ‘she’ and haar ‘her’; or, they use the feminine pronouns and state
in a note that these also refer to males. Again, this occurs only rarely.
The nominal gender system of Dutch is not static, but in a process of
change, and women are becoming more visible in the language. For many
speakers in the western and northern parts of the Netherlands, the mascu-
line/feminine distinction in nouns no longer exists, contrary to Dutch as
spoken in Belgium, where the distinction is fully alive (ANS 1997: 161). In
spoken Dutch, the “wrong” anaphoric pronouns are often used and referential
gender may override grammatical gender. In written Dutch, the “correct”
anaphoric pronouns are required.
Dutch 93
However, practices of spoken language have recently led to important

changes in Dutch dictionaries and grammars: Grammatical gender is gradually
being replaced by a natural gender system (cf. Schutz 1998: 294f). There are
changes in both the assignment of a noun to a gender class and the use of
pronouns. A number of personal nouns which in former editions of dictionar-
ies only had one grammatical gender (masculine), such as zieke ‘patient’ and
blinde ‘blind person’, are assigned two grammatical genders (masculine and
feminine) in recent editions (cf. Koenen 1986). Since the last edition of the
most authoritative dictionary of Dutch, Van Dale (1999), grammatical gender
no longer overrides referential gender. A whole section of the grammatical
description is devoted to the problem of incongruity of grammatical and
referential gender, and the solution suggested is that “designations for persons
such as arts (doctor), deskundige (expert) have a grammatical gender, but the
natural gender of the referent determines which personal pronouns have to be
used” (Van Dale 1999: xviii). When the gender of the referent is unclear,
however, the masculine pronouns hij, hem and zijn are used. The most authori-
tative grammar of the Netherlands, ANS (1997: 161), makes similar suggestions.
The change from grammatical to referential gender does not remove the
problem of gender-bias from the Dutch pronominal system, but it indicates a
development of the language in which women become more visible.
3.3 Terms of address and family names

In the area of address terms and family names, traditional practice is illustrated
by the following examples:
(21) Terms of address
Mevrouw: Term of address for married women and unmarried women
with a very high social or professional status (‘Mrs’)
Juffrouw: Term of address for unmarried women (‘Miss’)
Mijnheer: Term of address for men, regardless of marital and social
status (‘Mr’)
(22) Family names
Mevrouw Kool-Smit, Mevrouw Romein-Verschoor
Mevrouw Gerritsen, Mevrouw Jansen
Mijnheer Kool, Mijnheer Romein, Mijnheer Gerritsen, Mijnheer Jansen
Married women traditionally used their husband’s family name (e.g., Kool,
Romein) together with their own family name (e.g., Smit, Verschoor), or they
only used the family name of their husband (e.g., Gerritsen, Jansen). Married
men only used their own family name: Mijnheer Kool ‘Mr Kool’, Mijnheer
Romein, Mijnheer Gerritsen. Children received the family name of their father:
(23) Frank Jansen, child of Mr Jansen and Mrs Jansen-Bezema
3.3.1 Terms of address

Gender bias in terms of address has changed dramatically since 1975. There has
been an adaptation to the forms of address that were used for men. The
difference between a married and an unmarried woman – mevrouw versus
juffrouw – has disappeared completely. All women, irrespective of marital and
social status, are now mevrouw. The change began at the beginning of the
seventies in Amsterdam and spread all over the Netherlands within a decade,
starting in the towns of the western part and among the higher social classes.
Although no legal steps were taken to change traditional practices of addressing,
today there are no official forms any more in which the address form juffrouw
occurs. In 1993 a letter appeared in the feminist magazine Opzij in which an
unmarried woman of 80 complained that the term juffrouw was no longer in
use. She drew attention to the many questions she had to answer when she
introduced herself as mevrouw to the people in the home for the elderly where
she had been living for two years: When did your husband die? How many
children do you have?
It is remarkable that the disappearance of juffrouw went very smoothly.
The change had already taken place when the work groups on language sexism
in Dutch were founded. The groups did not have to struggle in order to
terminate its use.
Another example of the decrease of gender-biased language use is the
treatment of address terms in the dictionary by Van Dale. The edition of 1984
describes the correct ways of addressing the queen, a widow of noble birth, a
princess, a baroness, a countess, a nun, a sister, a lady, and eight other women
with a title. For these, an address form with vrouwe ‘madam’ is advised. All
other 67 (male) persons with a title should be addressed with a form of heer
‘sir’. This was all the more surprising since there were professions that even in
1984 were practised by women: lawyer, official, pharmacist, student, dentist.
The gender bias was solved in the edition of 1992, which states that women have
to be addressed with a form of vrouwe and men with a form of heer.
Dutch 95
3.3.2 Family names

For women, there has been a change from the use of the husband’s family name
to only their own maiden name. Before 1975, most women used both names
only in official documents, e.g., Kool-Smit, Romein-Verschoor. Some women did
it in more situations in order to express their own roots or facilitate the change
to re-using their maiden name. Most married women continued to use the
family name of their husband in everyday situations: mevrouw Gerritsen ‘Mrs
Gerritsen’. The change from using the husband’s name to the use of the
woman’s maiden name can only be described tentatively, since it has never been
investigated systematically. It has been conditioned by regional, social and age
factors. The change started just after World War II in Amsterdam among
married women in artist circles and spread to other groups during the sixties. At
the end of the seventies, nearly all women from the higher social classes who
were born after 1945 and lived in towns in the western part of the Netherlands
kept their maiden name or were going to take it again after having used their
husband’s name for some years. Some older women from the same group, but
born before 1945, also discarded the name of their husband, and used only their
own name; e.g., Joke Kool-Smit became Joke Smit. In many cases, the re-use of
the maiden name was related to divorce. Women born before 1930 kept the
name of their husband, even if they were divorced.
It was not until the beginning of the 1990s that married women could use
their maiden name in official forms. The change from using the family name of
the husband to using the maiden name is not complete yet, especially not
among the lower social classes, people living in rural areas and in the eastern
part of the Netherlands.
As was the case with the change from juffrouw to mevrouw, the change in
the use of family names proceeded rather smoothly. It was hardly an issue for
the work groups on language sexism. Nowadays, both possibilities are employed
side by side and there is no discussion about it. By using her maiden name a
woman can show that she is an emancipated woman. If she takes the name of
her husband, she indicates an adherence to more traditional norms.
Related to the use of the maiden name is the children’s family name. Until
1994, children received the father’s family name if the father and mother were
married. This was also the case when they were not married but the father had
officially recognised the child. Children with an unmarried mother and not
recognised by the father got the mother’s family name. In 1998, the law regard-
ing family names was changed and parents could choose between the mother’s
and the official father’s family name. The only condition was that children from
the same parents should have the same family name. Unfortunately, there are
no data yet about the choices made.5
3.4 Idiomatic expressions and stereotypical descriptions of women

and men
The way Dutch was used before the second feminist movement resulted in an
unequal linguistic treatment of women and men: the invisibility of women in
idiomatic expressions (cf. 24, 25), portrayal of women and men in stereotypical
gender roles (cf. 26, 27) and differences between women and men in how they
are described. In (28), Engelien Jansen is a woman, and C. Jansen is a man.
(24) Geen man over boord.
lit. no man over board
‘This is not a very big problem.’
(25) Een man een man, een woord een woord.
lit. a man a man, a word a word
‘You have to do what you say.’
(26) Zij heeft de boot gemist.
lit. she missed the boat
‘She did not marry.’
(27) Hij is de dans ontsprongen.
lit. he had a lucky escape
‘He did not marry.’
(28) Engelien Jansen, moeder van drie kinderen, en dr. C. Jansen zijn alle twee
hoogleraar.
‘Engelien Jansen, mother of three children, and Dr. C. Jansen are both
professors.’
Regarding idiomatic expressions in which women are either invisible or

portrayed in stereotypical gender-roles, the attitude of those people who
advocated a more gender-fair usage has been rather pragmatic. They advised
not to use such expressions or to use them also with the word vrouw ‘woman’
instead of man, or with ‘he’ instead of ‘she’ or ‘she’ instead of ‘he’ if required by
context and situation. A woman who wants to express that she intends to keep
a promise may use the expression ‘a woman a woman, a word a word’ instead
of (25). It is difficult to investigate whether this advice has been successful, but
it seems that the majority of the Dutch handle these expressions in a bias-free
Dutch 97
manner or at least try to do so.6 I have often observed that expressions such as
(24) and (25) slip out and are corrected immediately.
Regarding gender bias in the description of women and men, people with
a positive attitude towards a more gender-fair Dutch have demonstrated how
this sort of gender bias can be identified: Usually there is none if one can replace
male forms by female forms and vice versa. These suggestions have aroused a
keen interest and are discussed in prominent Dutch composition books. In
addition, differences in descriptions of women and men have been investigated
systematically in a number of genres: dictionaries, schoolbooks, television,
advertising, and newspapers (Brouwer 1991; Verbiest 1997, 1999). An indica-
tion of the change towards a more gender-fair usage is the fact that in the
prestigious dictionary of Van Dale, many sexist descriptions of words have
disappeared since the 10th printing (1976):
(29) a. Van Dale (1976)
Zij is aan de afwas.
‘She is washing the dishes.’
De meid heeft de afwas weer laten staan.
‘The maid has again not washed the dishes.’
b. Van Dale (1984)
Hij is aan de afwas.
‘He is washing the dishes.’
c. Van Dale (1992)
Hij heeft de afwas weer laten staan.
‘He has again not washed the dishes.’
To summarise, overt gender bias in Dutch has mostly disappeared since 1975.
Recent investigations of texts show that gender bias still occurs, although it is
far less noticeable.
3.5 Occupational terms: A case of Dutch language politics

As mentioned earlier, the beginning of the second feminist movement in the
Netherlands is marked by the appearance of Kool-Smit (1967), where it was
argued that the segregation between women’s and men’s professions had to be
removed. This required two major changes: First, it had to be made possible for
women to practise so-called men’s professions (e.g. carpenter) and for men to
practise so-called women’s professions (e.g. midwife), and second, since some
names for professions only referred to men and others only to women, new
professional titles had to be created.
Romein-Verschoor (1975) observed a tendency in Dutch towards gender-
specification of occupational titles; e.g., a female psychologist is called psycho-
loge and a male psycholoog. Furthermore, she shows that for some professions
only female-specific terms exist, e.g., verpleegster ‘nurse’, or vroedvrouw ‘mid-
wife’, while only male-specific terms exist for other professions, e.g., timmer-
man ‘carpenter’. In addition, she shows that when a presumably neutral term
such as arts ‘doctor’ is used, people only have a male person in mind.
The discussion about the problem of gender bias in professional names in
Dutch gained momentum when the law of Gelijke behandeling van mannen en
vrouwen bij de arbeid ‘Equal treatment for men and women regarding labour’
was established in 1980. In this law it was laid down that advertisements for the
recruitment of personnel had to be formulated in such a way that it was clear
that both women and men could apply. Advertisements that did not state this
clearly were legally forbidden. This law led to chaotic practices in the use of
names for professions, e.g., in job advertisements. Numerous solutions were
suggested, some organisations only added v/m, i.e. the abbreviation of vrouw
‘woman’/man ‘man’, to the professional term. This sometimes resulted in
amusing constructions such as (30), in which a term with a component that
clearly indicates a woman (-vrouw and -esse) got an addition that indicated that
men could also apply. By contrast, in (31) the names of the professions include
-man and -is, with additions that show that women could also apply:
(30) vroedvrouw (m/v) ‘midwife (m/f)’
secretaresse (m/v) ‘female secretary (m/f)’
(31) timmerman (m/v) ‘male carpenter (m/f)’
secretaris (m/v) ‘male secretary (m/f)’
Others tried to solve the problem through constructions such as (32), which is
a combination of the male/masculine suffix -eur and the female/feminine suffix
-ice, or a complicated combination of psycholoog ‘psychologist’ and psychologe
‘female psychologist’:
(32) directeur/trice ‘male/female director’
psycholo(o)g/e ‘male/female psychologist’
The Ministry for Social Affairs felt the need to intervene and set up a committee
called “Work Group Modification of Names for Professions” that was to write
Dutch 99
guidelines for the use of occupational terms which would address both women
and men. The work group consisted of members of the feminist movement
Man Vrouw Maatschappij and officials from the Ministry of Social Affairs. They
analysed the problem and published their directions in the brochure Gevraagd
which appeared in 1982 (Werkgroep 1982).
3.5.2 The recommendations and the public debate

On the basis of an analysis of Dutch professional titles, the “Work Group
Modification of Names for Professions” recommended the use of so-called
neutral terms as listed in (6) above. For stereotypically female (or male) terms
as in (5) they suggested new neutralising forms, for example timmer for
timmerman, huishoudelijke hulp for werkster or formations with neutral
suffixes such as -kracht and -wacht. The choice of the neutral names was
motivated as follows:
– In line with Romein-Verschoor (1975) the group stated that it was in principle
unimportant to know the gender of a person who practises a profession.
– On a social level, gender differentiation in professions would not be present
any more and many women would practise professions that were hitherto
mainly performed by men. As a consequence, the tendency to think only of
a man when perceiving the occupational term would disappear.
– On a linguistic level, if separate professional titles were introduced for
women and men, it was likely that a functional differentiation would occur
soon, the terms referring to women gaining lower social prestige than the
terms referring to men.
– When no neutral term was available, the term for men was chosen because
that was the more prestigious term.
– When neutral terms were used or masculine terms were already being used
for both men and women, it was not necessary to introduce new variants
such as hoogleerkracht ‘professor’, schrijfpersoon ‘author’.
The choice of neutral terms is in line with the change in the Dutch language,
which shows a decrease in number of productive suffixes referring to women
(for example -in, -es have disappeared) from the Middle Ages on.
The brochure Gevraagd (cf. Werkgroep 1982) and the guidelines resulted in
an avalanche of reactions in the media as well as in linguistic circles. The
discussion centred around three issues:
a.New formations such as timmer were considered ridiculous. Both lay-

persons and linguists were convinced that they would never gain a firm
foothold in Dutch.
b.Some linguists argued that the new terms were not neutral at all, but only
referred to men. Since not many women had practised professions such as
consul, dominee, minister, notaris, people would not immediately understand
that the respective job advertisements were also addressed to women. This
group, including van Alphen (1983) and Huisman (1985), advocated gender-
specific names for professions. Every name for a profession should have a
morphological equivalent for women. If such a name was not present in Dutch,
it had to be created, as in (33):
hooglerares hoogleraar ‘professor’
loodgietster loodgieter ‘plumber’
c.According to Adriaens (1982) there was indeed a decrease in the number of
productive suffixes referring to women in Dutch, but some suffixes (-e, -euse,
-trice) were used more frequently. Besides, he observed a tendency to introduce
the gender distinction also in the plural terms, which until now did not have
such a distinction: assistent and assistente both have the plural assistenten, but a
special plural for women assistentes was increasingly being used. Based on an
analysis of job advertisements in Dutch-speaking Belgium and in the Nether-
lands he believes that two conflicting tendencies can be observed: a tendency to
use gender-specific terms for professions and at the same time, a tendency to
use so-called neutral ones.
The “Work Group Language Sexism” of the VVAO largely supported the
advice of the “Work Group Modification of Names for Professions”. Regarding
pairs such as secretaresse and secretaris, they suggested using the terms as
indicators of different functions regardless of whether the person who practised
the function was male or female. They disapproved of forms like timmer, but
did not suggest any alternatives.
In the end, neither the guidelines of the “Work Group Modification of
Names for Professions” and the Ministry of Social Affairs, nor the proposals by
the group headed by Ingrid van Alphen and Joke Huisman, nor those of the
“Work Group Language Sexism” have become official guidelines. The discus-
sions in the first half of the 1980s were so controversial that no decisions could
be made.
Dutch 101
3.5.3 The manual for professional terms of the Dutch Language Union (2000)
In 1992, the Belgian Ministry of Employment and Labour, which also includes
the Department of the Emancipation of Society, commissioned the linguist
Patricia Niedzwiecki to develop guidelines for the feminisation of professional
terms in the three languages spoken in Belgium: French, Dutch and German.
Niedzwiecki recommended a long list of female-specific names for professions,
which was adopted by the government of the French-speaking part of Belgium.
It even issued a decree that gender-specific names for professions must be used
in official documents. However, the Dutch-speaking part of Belgium protested.
The linguist Willy van Langendonck argued that the proposal of Niedzwiecki
would lead to unwanted new formations such as ministerinne ‘female minister’,
and modification of words that had already been in use for a long time, e.g.,
lerarenopleiding ‘secondary teacher training course’ for which the variant
leraressenopleiding would be unwelcome. Besides, in his opinion the proposal
would end in confusion: The “best docente ‘female teacher’ of the year” is not
necessarily the same person as the “best docent ‘(male) teacher’ of the year”. The
Ministry also involved the linguist Johan de Caluwe, who warned against the
use of gender-specific terms for professions. Thus, no solution was found in the
Dutch-speaking part of Belgium.
In 1995, De Nederlandse Taalunie (Dutch Language Union) entered the
discussion (cf. Duynstee & Polak 1995). This is an official body of the govern-
ment of the Netherlands and the Dutch-speaking part of Belgium that prepares
and implements Dutch language and culture policy, with the objective to
advance the integration of the language and culture of the Netherlands and
Dutch-speaking Belgium. De Taalunie came to the conclusion that the choice
between the two alternatives, i.e. feminisation vs. neutralisation, was so charged
politically that no advice could be given at the time. The solution of the
Taalunie is – in line with its aim to develop a common knowledge and use of
Dutch – to make people aware of the problems and to suggest possible solu-
tions. It commissioned two linguists, the Dutch Ariane van Santen and the
Belgian Johan de Caluwe, to write a book entitled Gezocht: Functiebenamingen
(m/v). Wegwijzer voor vorming en gebruik van Nederlandse functiebenamingen
‘Requested: Professional titles (m/v). Manual for the formation and use of
Dutch names for professions’ (2001), which presents information about the
social and linguistic background of the issue and the various possibilities of
avoiding linguistic sexism. The handbook is merely meant to help people make
their own choice and to describe the political and linguistic implications of
alternatives.
4. The use of occupational titles in advertising: An empirical study
The question remains, then, of how professional titles are actually used.
Adriaens (1982) stated that only the future would show whether the tendency
to use neutralising terms or the tendency to use gender-specific terms would
win. In order to find an answer to this question, I analysed professional titles
in personnel advertisements in some of the 1999 issues of the same journals
that Snijders (1989) had examined in 1989, i.e. de Telegraaf and Intermediair.
Snijders only studied advertisements for professions for which a middle or
high education was required. She found 225 advertisements for such profes-
sions in one issue of Intermediair and two of de Telegraaf. In 1999, I found 573
of such advertisements in one issue of Intermediair and one of de Telegraaf.
The overall numerical difference between 1989 and 1999 is the result of an
increasing labour shortage during the last decade. It is difficult for companies
to find employees.
Table 2 shows the occurrence of four types of occupational titles in head-
ings of personnel advertisements in 1989 and 1999:
– use of masculine terms only (e.g. medewerker ‘co-worker’) in cases where a
feminine counterpart exists (medewerkster)
– use of “neutral” terms which refer to both women and men, e.g. arts (m)
‘doctor’
– use of feminine terms which only refer to women, e.g. secretaresse ‘female
secretary’
– use of mixed forms such as medewerk(st)er ‘(fe)male co-operator’
Table 2.Occupational titles in personnel advertisements in 1989 and 1999

1989 1999
N=225 advertisements N=573 advertisements
Masculine terms 104% 243%

Type: medewerker 46% 42%
“Neutral” terms 117% 323%

Type: arts (m) 52% 56%
Feminine terms 0% 6%

Type: secretaresse 0% 1%
Mixed forms 4% 1
Type: medewerk(st)er 1,7% 0,2%
Dutch 103
Table 2 shows that masculine and neutral forms predominate in the headings of
most advertisements, while feminine and mixed forms occur only rarely. This
is a strong indication that the tendency to use neutralising names will be the
preferred strategy. We even see an increase in the use of neutral names, but this
change is not yet statistically significant.
In Table 3, a distinction was made between the two types of neutral terms
for professions that were found in the data: Dutch terms and English terms. I
regarded terms as English if the word did not occur in Dutch. Terms that could
be both English and Dutch, for example accountant or supervisor, were consid-
ered Dutch, except when they occurred together with an English word, for
example people supervisor.
Table 3.Dutch and English “neutral” terms for professions

1989 1999
N=117 terms N=323 terms
Dutch [masculine terms] 47% 63%

hoofd ‘head’, arts ‘doctor’ 40% 19%
English 70% 260%
engineer, director 60% 81%
Table 3 shows a clear decrease in the use of Dutch neutral terms between
1989 and 1999 and an increase in the use of English terms. Not only were
English terms used more frequently in 1999, but also many new ones were
borrowed. In addition to the terms found by Snijders in 1989, we also found:
designer, developer, director, employee, floor broker, market maker, marketeer,
officer, professional, researcher, technician, telemarketeer, trader. It seems that the
use of English terms is one way of achieving gender-neutral expressions in
Dutch. This tendency is in line with the Anglicisation of Dutch society: English
is used in an increasing number of domains, e.g., in business, science, school
(cf. Nickerson 2000), and Dutch words are increasingly ousted by English
words, for example stomerij by dry cleaning, uitverkoop by sale (cf. Gerritsen &
Korzilius & Van Meurs & Gijsbers 2000).
The frequent occurrence of masculine and neutral professional titles in the
headings in advertisements indicates that neutralisation is increasingly used.
The question is whether the people who wrote these advertisements considered
the terms to be neutral, referring to both men and women, or whether they
indicated this in another manner. In the data, we found four different ways of
expressing that an advertisement was meant for both women and men:
– adding M/V (meaning Man/Vrouw ‘man/woman’) after the term
– using the pronouns hij/zij ‘he/she’ or zijn/haar ‘his/her’ as anaphoric
pronouns in the text
– adding a female-specific term in the text, e.g. kandidate ‘female candidate’
(a feminine term)
– explicitly asking women to apply.
Table 4 shows how often these strategies occur in advertisements with mascu-
line or neutral headings.
Table 4.Explicitly inclusive expressions

1989 1989 1989 1999 1999 1999
Masc. “Neutral” Total Masc. “Neutral” Total
terms terms terms terms
Addition 80% 77% 157% 109% 90% 200%

of m/v 77% 66% 71% 45% 28% 35%
hij/zij or zijn/haar No data No data No data 26% 56% 82%

11% 17% 14%
Addition of fe- No data No data No data 1% No data 1%

male-specific 0,4% 0.2%
term
Explicit No data No data No data 21% 9% 30%
mention 9% 3% 5%
Table 4 shows an enormous decrease in the use of the addition of M/V

between 1989 and 1999. In 1999 only 35% of the masculine and neutral terms
had an addition M/V, whereas this was the case in 71% in 1989. This is an
indication that the masculine and the neutralising terms are increasingly
considered to refer to both women and men. With regard to the other possibili-
ties of explicitly making reference to women, we only have the data for 1999,
which show that the anaphoric pronouns hij/zij occur relatively frequently, that
is to say, in 14% of the cases, but it is not very often that it is stated in the
description of the candidate that a woman is meant. Moreover, women are not
often asked to apply explicitly. In 58% of all advertisements with a masculine or
neutral heading (N = 566) of the 1999 corpus we did not find any of the four
Dutch 105
strategies to indicate that the advertisement was also meant for women. This
could be interpreted as a sign that the masculine and neutralising terms are
considered to be neutral. The change from grammatical to referential gender in
nouns such as arts ‘doctor’, or blinde ‘blind person’ supports this interpretation.
However, we doubt whether this is indeed the case. Psycholinguistic studies in
the line of Braun & Gottburgsen & Sczesny & Stahlberg (1998) and Braun
(vol. I) are needed to discover whether masculine professional titles such as
medewerker ‘co-operator’ and neutralising titles such as arts ‘doctor’ are indeed
perceived as referring to both women and men.
5. Conclusion
Obviously, there has been a decrease in the use of gender-biased language in the
Netherlands since 1975. From a language-political point of view it is amazing
that so many changes have taken place in such a short time. First of all, it shows
that if changes in society require changes in language, these changes can be
effectuated quickly, even on a grammatical level, such as the switch from a
grammatical to a natural gender system. Secondly, the change towards more
gender-fair usage in the Netherlands shows that the introduction of legal steps,
such as the law which forbids gender-biased advertising for the recruitment of
personnel, may have a considerable impact on language use and language
change. Last but not least, the changes in the Dutch language show that it is
worth the effort to fight against gender-bias in language use.
Notes
1. Additional reference works of Dutch are: Shetter (1994), Donaldson (1981, 1983), Martin
& Tops (1984).
2. Except for some older people, speakers of Standard Dutch as spoken in the Netherlands
do not know which nouns are masculine and which are feminine, which affects their choice
of pronouns. Speakers of Standard Dutch as spoken in Belgium (also called Flemish) can still
distinguish between grammatical genders (cf. ANS 1997: 234). With regard to names for
professions, the Dutch mainly use the same terms for both women and men, but the Flemish
use different terms (cf. Section 3.5.3).
3. The following discussion is based on Donaldson (1983:58ff), Kooij (1987), and ANS (1997).
4. The term domina meaning ‘female clergyman’ is archaic; today, the word means ‘female
lover’.
5. Officials of the registration of birth office told us in the summer of 1999 that the family
name of the mother is seldom chosen.
6. I do, of course, not know whether people do so only when I am present – they know that
I am a supporter of Dutch that is free of gender bias – or whether they habitually correct their
sexist language use in other situations as well.
References
Adriaens, Geert. 1982. “Vrouwelijke beroepsnamen in evolutie [Change in professional titles

for women].” Forum der Letteren 23: 1–17.
ANS = Algemene Nederlandse Spraakkunst [General Dutch grammar]. 1997. Groningen:
Martinus Nijhoff.
Braun, Friederike & Anja Gottburgsen & Sabine Sczesny & Dagmar Stahlberg. 1998.
“Können Geophysiker Frauen sein? Generische Personenbezeichnungen im Deutschen.”
Zeitschrift für Germanistische Linguistik 23: 265–283.
Brouwer, Dédé. 1980. “Taal en Sekse als probleem voor de taalwetenschap [Language and
gender as a problem for linguistics].” Wetenschap en Samenleving 2: 105–115.
Brouwer, Dédé. 1985. “Hoe vrouwelijk is een hoofd [How feminine is a head]?” De Gids
148: 105–115.
Brouwer, Dédé. 1991. Vrouwentaal. Feiten en verzinsels [Women’s language. Facts and
figments]. Bloemendaal: Aramith uitgevers.
de Caluwe, Johan & Ariane van Santen. 2001. Gezocht: Functiebenamingen (m/v). Wegwijzer
voor vorming en gebruik van Nederlandse functiebenamingen [Requested: Professional
titles (m/v). Manual for the formation and use of Dutch names for professions]. The
Hague: SDU Uitgevers.
de Vries, Jan W. & Roland Willemyns & Peter Burger. 1993. Het verhaal van een taal. Negen
eeuwen Nederlands [The story of a language. Nine centuries of Dutch]. Amsterdam:
Uitgeverij Prometheus.
Donaldson, Bruce C. 1981. Dutch reference grammar. Leiden: Martinus Nijhoff.
Donaldson, Bruce C. 1983. Dutch: A linguistic history of Holland and Belgium. Leiden:
Martinus Nijhoff.
Duynstee, Rob & Charlotte Polak. 1995. “Discussie over vrouwelijke beroepsnamen laait weer
op [Discussion over professional titles for women flares up again].” Taalschrift 14: 14–16.
Gerritsen, Marinel & Hubert Korzilius & Frank Van Meurs & Inge Gijsbers. 2000. “English
in commercials on Dutch Television: Frequently used, but not understood and not
appreciated.” Journal of Advertising Research 40: 17–35.
Huisman, Joke. 1985. “Over verschil in gelijkheid en gelijkheid in verschil [About difference
in equality and equality in difference].” De Gids 148: 777–783.
Koenen, M. J. 1986. Wolters’ woordenboek Eigentijds Nederlands. Grote Koenen (1986)
[Wolters’ dictionary of current Dutch. Large Koenen]. Groningen: Wolters-Noordhoff.
Kooij, Jan G. 1987. “Dutch.” In The world’s major languages, ed. Bernard Comrie. New York:
Oxford University Press, 139–156.
Dutch 107
Kool-Smit, Joke E. 1967. “Het onbehagen bij de vrouw [The discontentment of women].” De
Gids 130: 267–282.
Lakoff, Robin. 1975. Language and woman’s place. New York: Harper & Row.
Martin, Willy & Guy A. J. Tops. 1984. Van Dale groot woordenboek Engels–Nederlands and
Nederlands–Engels [Van Dale large dictionary English–Dutch and Dutch–English].
Utrecht: Van Dale Lexicographie.
Nickerson, Catherine. 2000. Playing the corporate language game. Amsterdam: Rodopi.
Renkema, Jan. 1995. Schrijfwijzer [Advice for writing]. The Hague: SDU.
Romein-Verschoor, Annie. 1975. “Over taal en seks, seksisme en emancipatie [About
language and gender, sexism and emancipation].” De Gids 138: 3–36.
Romein-Verschoor, Annie. 1995 [1977]. “Over taal en seksisme [About language and
sexism].” In Een zee van golven, eds. Agnes Verbiest & Anne Van Zande. Nijmegen:
VITA, 11–15.
Schutz, Rik. 1998. “Woordgeslacht en sekse [Grammatical gender and sex].” Onze Taal 11:
294–295.
Shetter, William Z. 1994. Dutch: An essential grammar. London: Routledge.
Snijders, Gina. 1989. Bent u die kandidaat (M/V)? Een onderzoek naar aanspreekvormen in
personeelsadvertenties [Are you the candidate (M/F)? An investigation of forms of
address in job advertisements]. Leiden: Rijksuniversiteit Leiden.
Van Alphen, Ingrid. 1983. “Een vrouw een vrouw, een woord een woord. Over de gelijke
behandeling van vrouwen en mannen en de konsekwenties daarvan voor beroeps-
benamingen in het Nederlands [A woman a woman, a word a word. About the equal
treatment of women and men and the consequences for professional titles in Dutch].”
Tijdschrift voor Vrouwenstudies 14: 306–315.
Van Dale. 1976. Groot Woordenboek der Nederlandse Taal [Large dictionary of the Dutch
language]. 10th ed. The Hague: Martinus Nijhoff.
language]. 11th ed. The Hague: Martinus Nijhoff.
language]. 12th. ed. Utrecht: Van Dale Lexicografie.
language]. 13th ed. Utrecht: Van Dale Lexicografie.
Van Gessel, Han & Jan Kees Hulsbosch & Henk Huurdeman & Bas van Kleef & Kees Los &
Bert Vuijsje. 1992. De Volkskrant stijlboek [The Volkskrant style guide]. Amsterdam:
De Volkskrant.
Verbiest, Agnes. 1994. “Onderzoek naar taalseksisme: Theorie en toepassing [Research in
sexist language: Theory and application].” Vaktijdschrift voor Volwasseneducatie en
Sociaal-Cultureel werk 9: 291–305.
Verbiest, Agnes. 1997. De oorbellen van de minister. Taal en denken over vrouwen [The earrings
of the minister. Language and thinking about women]. Amsterdam: Uitgeverij Contact.
Verbiest, Agnes. 1999. Zaken zijn zaken. Taal en de kwaliteit van het beleid [Business is
business. Language and the quality of government policies]. The Hague: Ministerie van
Sociale Zaken en Werkgelegenheid.
Werkgroep Wijziging Beroepsbenaming. 1982. Gevraagd. Aanzet tot een discussie over het
wijzigen van den beroepsbenaming in het kader van de wet gelijke behandeling mannen en
</TARGET "ger">
vrouwen [Demanded. Initial motivation for a discussion of the modification of profes-

sional titles as related to the law of equal treatment of men and women]. Amsterdam:
Uitgave Ministerie von Sociale Zaken en Werkgelegenheid en de aktiegroep Man Vrouw
Maatschappij.
Woordenlijst Nederlandse Taal [Wordlist of the Dutch language]. 1995. The Hague: Uit-
geverij Areopagus.
<TARGET "eng" DOCINFO
AUTHOR "Mila Engelberg"
TITLE "The communication of gender in Finnish"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
finnish
The communication of gender in Finnish
Mila Engelberg
University of Helsinki, Finland
1. Introduction
2. Women’s position in Finland
3. Research on gender in Finnish
4. Gender in Finnish
4.1 Personal nouns: Referential and morphological aspects
4.2 Male generics
4.3 Female generics
5. The interpretation of male generics: Empirical evidence
6. The empirical investigation of covert male bias
6.1 The prototypical human being
6.2 Ihminen as ‘woman’
7. Gender in Finnish proverbs
8. Gendered terms of abuse
9. Female/male discourse
10. Language reform
11. Conclusion
Notes
References
1. Introduction
Finnish (suomi or suomen kieli) is a Finno-Ugric language spoken by about five

million people in Finland. In addition, there are some 300,000 speakers of
Finnish in Sweden, 12,000 in Northern Norway, as well as Finnish-speaking
minorities in Eastern Karelia, Ingria, in the USA, Australia and Canada.
Finland belonged to Sweden until it became an autonomous Grand Duchy
within Russia (from 1809–1917). Finland declared its independence in 1917.
110 Mila Engelberg
In 1863, Finnish was given official and equal status with Swedish, and today
Finland has two national languages, Finnish and Swedish. The history of written
Finnish is relatively short, the first books published in Finnish dating from the
beginning of the 16th century. There are eight regional dialects of Finnish, the
standard language being primarily based on the southwestern dialects.
Finnish is an agglutinating language with various markers and clitics,
fifteen cases and about 150–200 derivational affixes which are typically added
to the word stem. The word in (1) illustrates the morphological complexity
and also the morphotactics of Finnish: root – derivational affixes – inflectional
affixes – clitics:
(1) juo -pu -isi -vat -ko
drink -der -cond -3pl -question.prt
‘would they become drunk?’
Another characteristic feature of the language is the morphophonemic alterna-

tion of various kinds caused, among other things, by vowel harmony. The
unmarked word order is Subject-Verb-Object-Adverbial. Finno-Ugric languag-
es have had repeated language contact with Indo-European languages. Basic
vocabulary differs from that of Indo-European languages, but due to historical
reasons Finnish has incorporated hundreds of loanwords and loan translations
mainly from Swedish (cf. Karlsson 1983: 10f; Lehikoinen & Kiuru 1991: 1–10;
Sulkala & Karjalainen 1992: xiv f).1
2. Women’s position in Finland
Finnish society is usually considered to be relatively gender-equal. It is often

pointed out that Finnish women were the first in Europe to receive the right to
vote in 1906. In addition, the first female parliament members in the world
were elected in Finland in 1907. Women hold 37% of the seats in the parlia-
ment elected in 1999. Seven cabinet ministers out of 18 are women. The first
female president of Finland was elected in 2000. At the same time, such
political decision-making bodies as the parliament are losing their power in
Finland. It seems that, once again, “power escapes women”.
Education as well as the labour market are gender-segregated. Most women
are wage earners but still carry a double burden, performing 65% of the
household chores. The same level of education does not guarantee women
equal positions with men in the labour market. Sensitive issues such as sexual
Finnish 111
harassment and violence against women became public issues in the late 1980s.
The economic crisis in the early 1990s weakened the pro-women Finnish
welfare state, and feminists have begun to talk about a backlash against women
(cf. Husu & Niemelä 1993).
3. Research on gender in Finnish
In Finland, gender mainly became an object of linguistic analysis in the early 1980s
as a factor in sociolinguistic research on speech variation (cf. Hakulinen & Laitinen
1993). For instance, Nuolijärvi (1986) in her research on linguistic adaptation
of those speakers moving to Helsinki from southern Ostrobothnia and northern
Savo found that women were more colloquial than men, while men used both
more standard language and more dialectal features than women. Research has
also brought out the interdependence of gender and other extralinguistic factors
such as age, level of education and occupation (e.g. Mielikäinen 1988). During
the 1990s, researchers became interested in a more interactional approach and
the method of conversation analysis (see Section 9 below).
The absence of grammatical gender in Finnish (cf. Section 4.1) has contrib-
uted to the perception of Finnish as a comparatively gender-neutral language.
This may explain why the issue of gendered or sexist language has received less
attention in Finnish than in French or German, for example. In a recent study,
Kangas (1997) investigated the otherness of woman in Beauvoirian terms in
written language use and found that both female and male university students
refer to women more often than to men in relation to other people: Expressions
such as äiti ‘mother’ or Leila-täti ‘aunt Leila’ were more frequent than for
example poikakaveri ‘boyfriend’ or Heikki-setä ‘uncle Heikki’. Engelberg
(1998a) examined gendered occupational terms in Finnish (e.g., bingoemäntä
‘bingo hostess’) from the directory of occupations published by the Central
Statistical Office of Finland in 1950, 1970 and 1990. One of the major findings
was the infrequency of compounds with nainen ‘woman’ in all three directories
in comparison to hundreds of titles with mies ‘man’. One explanation for this
asymmetry may be the argument that a greater productivity of nainen in
occupational titles would highlight the complementary opposition of nainen/
mies and narrow the meaning of many titles with mies, making them more
male-specific, and thus weakening their apparent genericity.
112 Mila Engelberg
In the following I shall discuss gender in the Finnish language system, as

well as various features of gendered language as these are manifested in Finnish,
including proverbs and terms of abuse. I will focus on lexical androcentricity in
the so-called male generics2 and in the covert male bias of terms with human
reference.
4. Gender in Finnish
4.1 Personal nouns: Referential and morphological aspects

Like other Finno-Ugric languages, Finnish lacks grammatical gender (on
Turkish, cf. Braun, vol. I). The personal pronoun system does not distinguish
female and male forms. The nominative forms of personal pronouns are shown
in Table 1.
Table 1.Finnish personal pronouns (nominative form)

Singular
Person 1 minä ‘I’
2 sinä; Te ‘you; you-formal’
3 hän (se, tämä, tuo) 3 ‘he/she (it, this, that)’
Plural
Person 1 me ‘we’
2 te ‘you’
3 he (ne, nämä, nuo) ‘they (they-inanimate, these, those)’
As Table 1 illustrates, not even 3rd person singular pronouns are differenti-
ated for gender. The semantic feature that distinguishes hän ‘she/he’ from se,
tämä, tuo ‘it, this, that’ is [±animate].
As Braun (1995: 11) points out, the Finnish possessive expression lääkä-
rinnsä ‘doctor-3poss’ can denote ‘her (female) doctor/her (male) doctor/his
(female) doctor/his (male) doctor’. Unlike man in English or homme in
French, the Finnish word for ‘man, male’, mies, denotes male gender only.
However, ihminen ‘human [being], man’ and ihmiskunta ‘mankind, humanity’
are gender-neutral. ‘Man is mortal’ is translated as Ihminen on kuolevainen, not
*Mies on kuolevainen (cf. German Mann ‘male person, man’ vs. Mensch
‘human being’).
In Finnish, referential gender can be indicated by nouns with lexical gender
as well as by compounding and to some extent by derivation. In addition to
such frequently used words as nainen ‘woman’, tyttö ‘girl’ and mies ‘man, male’,
Finnish 113
human nouns with lexical gender typically belong to the domain of kinship
terminology. Certain kinship terms such as leski ‘widow, widower’ can refer to
both women and men. Some of the basic kinship terms are shown in Table 2.
Gendered terms of address have become rare during the last two or three
decades, although they can still be used in polite speech. A man is addressed
with the title herra ‘Mr, sir’, e.g., Herra Presidentti ‘Mr President’, anteeksi herra
‘excuse me, sir’. Terms with female reference comment on the woman’s marital
status: neiti ‘Miss’ and rouva ‘Madam, Mrs’. The president elected in 2000 is
addressed as Rouva Presidentti ‘Madam President’.
Referential gender can also be specified by modifiers such as nais(puolinen)
‘female’ or mies(puolinen) ‘male’ in compounds:
(2) liikenainen ‘business woman’
laivapoika ‘shipboy’
naispoliisi ‘woman police officer’
miesmalli ‘male model’
naispuolinen väestö ‘female population’
miespuolinen jälkeläinen ‘male offspring’, leg. ‘male issue’
Finnish has four female suffixes, none of which is fully productive (Kyrölä 1990,
Vesikansa 1978). The affix -kkO (O is realized as ö or o according to vowel
harmony) is of Finnish origin and can be used in the formation of nouns with
female reference, e.g., sisäkkö (from sisä ‘interior’) ‘parlourmaid’, venakko (from
Venäjä ‘Russia’) ‘a Russian woman’ and a recent, jocular formation sihteeri-kkö
(from sihteeri ‘secretary’) ‘secretary’.4 The suffix -tAr (A stands for ä or a)
originates from the word tytär ‘daughter’. Nouns with this suffix often refer to
dignitaries or inhabitants of certain areas, e.g., kreivi-tär ‘countess’ and pariisi-
tar ‘Parisienne’; lexicalised forms include kauno-tar ‘beauty, bell’ and rakastaja-
tar ‘mistress’. Occupational titles such as opettaja-tar ‘female teacher’ have not
been in official use since the 1970s (Engelberg 1998a).5 Titles with -tAr can
often be found in the sports pages, e.g., hiihtäjä-tär ‘female skier’. Folk poetry
Table 2.Finnish kinship terms

Female Male
äiti ‘mother’ isä ‘father’
sisar, sisko ‘sister’ veli ‘brother’
tytär ‘daughter’ poika ‘son, boy’
anoppi ‘mother-in-law’ appi ‘father-in-law’
täti ‘aunt’ (mother’s/father’s sister) setä ‘father’s brother’
eno ‘mother’s brother’
114 Mila Engelberg
also knows such mythical female creatures as luonno-tar ‘nature’ (from luonto
‘nature-FEM’). The derivatives with -skA and -nna are of Swedish origin and
have now become archaic or colloquial. Both suffixes were previously used in
reference to civil servants’ wives, e.g. professor-ska ‘professor’s wife’ and eversti-
nna ‘colonel’s wife’; -skA also appears in such female workers’ titles as huusho-
ller-ska (coll.) ‘house keeper’ and in married women’s last names, e.g., Aalto-ska
(coll.) ‘wife of Aaltonen’.
4.2 Male generics

The perception of people as male (or male as people) is made explicit in male
generics, i.e. generic expressions which contain an element with the semantic
specification [male], in particular -mies ‘man’ or -isä ‘father’. Finnish male
generics are compounds, derivations and idioms (cf. Karlsson 1974). “Generic”
human nouns in particular are often described as a logical reflection of (origi-
nally) male-dominated reference groups. Many titles do in fact refer to tradi-
tionally male activities in the Finnish culture, as in (3). Note that hyphens are
used to indicate morphological structure; they do not correspond to ortho-
graphic conventions.
(3) metsä-mies lit. ‘forest-man’ ‘hunter, sportsman’
tupakka-mies lit. ‘cigarette-man’ ‘smoker’
However, male generics are not solely a product of extralinguistic gender
segregation. Expressions such as (4) rather reflect an androcentric conceptual-
ization of reality:
(4) joka-mies lit. ‘every-man’ ‘everybody’
joka-miehen oikeudet lit. ‘every-man’s rights’ ‘public right of access’
esi-isä lit. ‘fore-father’ ‘ancestor’
viikate-mies lit. ‘scythe-man’ ‘the Grim Reaper, Death’
Compounds with non-human reference include isäntä-maa ‘host country’ and
veljes-sota (< veli ‘brother’) ‘brother-war, civil war’.
Male generics can also be formed by derivation:
(5) isäntä ‘host’ + -Oi > verb isännöidä ‘act as a host’
mies ‘man’ +-istO > collective noun miehistö ‘the crew, the men’
Common idiomatic expressions are miesmuistiin lit. ‘in man’s memory, in
living memory’ and olla oma herransa ‘be one’s own master’. Idioms and
phrases referring to men often have a positive connotation, e.g.:6
Finnish 115
(6) Juho toimi miehe-n lailla tiuka-ssa tilantee-ssa.

Juho act.past.3sg man-gen like tough-iness situation-iness
‘Juho acted like a man in a tough situation.’
(From a TV program in 1998; refers to a seven-year-old boy who called
the emergency number after a fire had broken out at his home.)
Politician Marjatta Väänänen writes in her memoirs about the time when she
was acting as the minister of cultural affairs and in 1974 rejected a political
attack from leftist cultural radicals. A male newspaper columnist commented
that the minister had acted like a man. Another male journalist wrote: “There
is at least one man in the Finnish government: Marjatta Väänänen” (Väänänen
1996: 144f). During the 1994 presidential election campaign the candidate
Elisabeth Rehn was praised by a male radio reporter: “Rehn oli ainoa, joka
osoitti miehistä suoraselkäisyyttä” ‘Rehn was the only one who showed manly
uprightness’. Nowadays a woman can also be given credit for being hyvä jätkä
‘a good guy’.
4.3 Female generics

Female generic expressions form a more homogeneous group than male generic
forms. They are usually compounds referring to genealogical relations, e.g.,
sisaryhtiö ‘sister company’ (cf. Karlsson 1974: 27). According to Karlsson, this
may reflect associations of a female fertility cult which in other languages is
sometimes manifested in some grammatically feminine nouns denoting the
earth or plants; cf. German (die) Erde, Latin terra. Other Finnish examples are
tytärsolu ‘daughter cell’, äidinkieli ‘mother tongue’ and Suomi-neito ‘maiden
Finland’, a metaphor for Finland. Female generics usually do not have human
reference. One of the exceptions is sisarus ‘sibling’ or sisarukset ‘siblings’ (from
sisar ‘sister’). According to my experience, sisarukset is not readily used when
referring to brothers only (see also Rautala 1988: 33). Another female generic
expression with human reference is morsius-pari ‘bridal-pair; bride and
bridegroom’, although it may be giving way to hääpari ‘wedding-couple’, which
is clearly more common on congratulation cards, for instance.
Female occupational terms are not used generically. The first male flight
attendants did not inherit the existing title lentoemäntä ‘flight-hostess, stew-
ardess’, but were called stuertti ‘steward’ from the beginning. Similarly, male
116 Mila Engelberg
workers were not called karja-kko (lit. ‘cattle-female’) ‘cattle maid’ or meijeri-
kkö (lit. ‘dairy-female’) ‘dairy maid’ but karjanhoitaja ‘cattle tender’ and
meijeristi ‘dairy man’, respectively.
5. The interpretation of male generics: Empirical evidence
It seems that speakers are aware of the ambiguity and the limited generic
potential of male generics. A foundation for legal research announces that their
research grants for the year 1993 are available for lakimiehelle tai naiselle7 lit.
‘for a law-man or a law-woman’. Had the announcement used the morpholog-
ically gender-neutral synonym juristi, a specification such as this would hardly
have been necessary. Unless otherwise stated, research grants are available for
male as well as female applicants. To give another example, an ad for a skiing
magazine uses the following phrase: Autamme miestä mäessä. Ja naista ‘We will
help a man up the hill. And a woman’.
Empirical research on male generics in Finnish is, however, only in its
beginnings. In Engelberg (1993, 1995; cf. also Braun 1997b: 41–43), 76 female
and 74 male university students (mean age 23.5 years) were presented with
sentences containing male generic forms, e.g.:
(7) Tavallinen kadunmies ei juuri jaksa seurata budjettikeskustelua.
‘The man in the street doesn’t really have the energy to follow the discus-
sion about the budget.’
Informants rated these sentences as referring to men significantly more often

than (otherwise identical) sentences containing morphologically neutral
expressions, e.g.:
(8) Tavallinen kadunkulkija …
‘The average person …’
The target sentences included one pair of verbs, veljeillä ‘to fraternize’ vs. kavee-
rata ‘to pal around with somebody’, and two pairs of idioms, as in (9):
(9) a. joutua maksu-miehe-ksi
end.up pay-man-trans
b. joutua maksaja-ksi
end.up payer-trans
‘to end up paying’
Finnish 117
Male participants rated “generic” expressions as referring to men significantly

more often than female participants did. The participants in the male generic
condition gave a total of 949 responses; the referents were perceived as male in
56%, as gender-neutral in 42% and as female in less than 3% of the responses.
Two further studies on male generics were designed to elicit unmonitored
responses from participants (for a detailed discussion see Engelberg 1998b). All
subjects had Finnish as their native tongue and participated voluntarily.
Thirty-eight female and sixty male students from two high schools specializing
in arts education in Espoo participated in the first study. They were presented
with five sentences referring to workers. The sentences had been adapted from
the collective bargaining contract for government officials and the correspond-
ing contract for municipal employees. The texts consisted of various concrete
terms and conditions of employment: sick leave, coffee breaks, uniforms,
medical examination, or lecture fee for a worker giving a lecture. Two of the
sentences were fillers and contained the word työntekijä ‘worker’. The three
other texts were target sentences and referred to the ‘worker’ with the generic
male title virkamies ‘office-man, civil servant’.8 The participants were instruct-
ed to illustrate each sentence and make a drawing of the worker mentioned in
the sentence, using the ideas that first came to their minds. In addition, they
were instructed to give a first name to each character. Of the 291 responses
codable by gender, 80% described virkamies as male. Boys interpreted virkam-
ies significantly more often as male than as female in all three paragraphs, girls
only in one of the three paragraphs. Overall, 65% of the characters produced
by girls and 90% of those created by boys were men.9
The second experiment was carried out in eleven schools in Helsinki.
Participants were 275 high school students (175 women, 100 men; mean age
16.3) and 231 students from the upper level of a comprehensive school (126
girls, 105 boys; mean age 12.8).
The participants were presented with a form displaying 22 sentences which
referred to workers in various occupations. Three of the sentences were target
sentences, including a generic male title in the experimental group and a
morphologically neutral synonym in the control group, e.g.:
(10) Myymälä-esimies ‘shop-foreman’ / Myymälä-n-hoitaja ‘shop-gen-tender’
tilasi uuden erän Leviksiä.
‘The shop manager ordered a new lot of Levi’s.’
The other target sentence pairs included the titles ammattiliiton luottamusmies
‘trade union’s trust-man’/ammattiliiton luottamushenkilö ‘trade union’s trust-
118 Mila Engelberg
person, trustee’ and lakimies ‘lawman’/juristi ‘lawyer’. All occupational titles

appeared in subject position.
The filler sentences referred to occupations in which neither women nor
men represent less than 40% of the employees (Central Statistical Office of
Finland 1996, unpublished statistics) and/or which have prominent female as
well as male figures (e.g., Uutistenlukija luetteli pääuutisaiheet ‘The newscaster
read the main headlines’).
The participants were instructed to try to picture in their minds the person
referred to in the sentence and then give this person a first name. The purpose
of the study was said to be the analysis of the connotations of first names.
Myymäläesimies ‘shop-foreman’ elicited significantly more men’s names
than myymälänhoitaja ‘shop-tender’ from male as well as female participants.
The referent of luottamusmies ‘trust-man’ was interpreted as male significantly
more often than the referent of luottamushenkilö ‘trust-person’ by female, but
not by male participants. The sentence with lakimies ‘lawman’ did not produce
significantly more men’s names than the corresponding sentence with juristi
‘lawyer’ (but the difference did approach significance among comprehensive
school girls).10
The data from the second experiment show how the semantic gender
neutrality or gender bias of the neutral equivalents as opposed to the male
generics may vary.11 At the same time, however, the unmonitored responses in
the study replicate the male bias in the interpretation of male generic agent
nouns: 85–89% of all names given in response to the generic male sentences
were men’s names. In conclusion, none of the three studies above supports the
description of male generic expressions in Finnish as indeed generic and
unmarked.
The interpretation of male generics in the first and second study does not
necessarily reflect the distribution of women and men in the corresponding
referent groups. For example, women have had the majority in the referent
group denoted by virkamies, the public sector employees, at least since the
beginning of the 1980s.12
6. The empirical investigation of covert male bias
Researchers on gendered language have found that not only generic mascu-
line/male expressions, but also morphologically gender-neutral human nouns
(which contain no gender-specific element) can have a male connotation (e.g.,
Finnish 119
Braun 1995, 1997a, cf. also the contribution by Braun, vol. I; Hamilton 1991).
I shall refer to this phenomenon as covert male bias. A reanalysis of the gender-
neutral sentences and their perceived referents in Engelberg (1993) reveals a
covert male bias in the participants’ reactions. The referents of the “gender-
neutral” sentences were judged as male in 30% and as female in 6% of the
responses. The referent was perceived as female equally often by female and
male participants. The actual percentage of women and men in the referent
groups (as far as this was possible to specify) did not explain the participants’
responses. For example, the referent of the word kolleega ‘colleague’ in the
sentence Tapasitko eilen ruotsalaisen kolleegasi? ‘Did you meet your Swedish
colleague yesterday?’ was perceived as male in 37% and as female in 2% of the
responses.
Covertly male language use is not uncommon in the media, for example. In
some instances, the androcentricity is blatant, in others the message is more
hidden. Consider examples (11–14) below:
(11) Puna-armeija-n nais-komissaari Klavdia Vavilova […]
Red-Army-gen female-commissar Klavdia Vavilova […]
huomaa tul-lee-nsa raskaa-ksi.
notice.3sg come-part-poss pregnant-trns
‘Female Red Army commissar Klavdia Vavilova […] finds out that she is
pregnant.’
The woman’s title komissaari in example (11) clearly denotes a male-dominated

group and contains a male connotation. The need to specify the female refer-
ent’s gender with the modifier nais- ‘female’ has, however, led to a pleonastic
expression. The referent’s gender is evident from the proper name as well as
from her pregnancy.
Unlike in the case above, covert masculinity in the following example
cannot be explained by extralinguistic factors:
(12) Läheise-ssä kala-ravintola-ssa pöydä-t notku-vat
nearby-iness fish-restaurant-iness table-pl bend/give.way-3pl
katkaravu-i-sta […] Asiakkaa-t o-vat kiinalais-i-a,
shrimp-pl-elat […] customer-pl be-3pl chinese-pl-partt
hyvin pukeutu-ne-i-ta perhee-n-is-i-ä
well dress-part-pl-partt family-gen-father-pl-partt
120 Mila Engelberg
vaimo-i-neen ja laps-i-neen.
wife-pl-com and child-pl-com
‘The tables in a nearby fish restaurant are filled with shrimps […] The
customers are Chinese, well-dressed family men with their wives and chil-
dren.’
Example (13) demonstrates that the prototypical meaning of suomalainen ‘a

Finn’ also has a male bias:
(13) “Suomalaise-lle alkoholi tarkoittaa pari-a
Finn-allat alcohol mean-3sg couple-partt
pullo-a keski-olut-ta sauna-n jälke-en.
bottle-partt middle-beer-partt sauna-gen trace-illat
Ne nauti-taan koto-na vaimo-n kanssa
they enjoy-pass home-ess wife-gen with
lotto-tuloks-i-a odotel-le-ssa”, Piispa sano-o.
lotto-result-pl-partt wait-inf-iness Piispa say-3sg
“For a Finn alcohol means a couple of bottles of beer after the sauna.
They are consumed at home together with the wife while waiting for the
Lotto results”, says Piispa.
During the presidential election campaign in 1994, a male TV-reporter asked

the pro-feminist candidate Elisabeth Rehn: “Oletteko te nainen vai ihminen?”
‘Are you a woman or a human being?’ The meanings of ihminen and nainen are
also differentiated in example (14):
(14) Highway-n ihmis- ja nais-suhtee-t o-vat
Highway-gen human- and woman-relation-pl be-3pl
pala-nee-t tietenkin karre-lle.
burn-part-pl of.course snuff-allat
‘Highway’s relationships with people and women have of course stranded.’
(Highway is a personal name.)
In the following I shall report on the major findings from two empirical studies
on covert gender in morphologically neutral human nouns in Finnish (see
Engelberg 2001 for a detailed discussion).
6.1 The prototypical human being

109 high school students (74 women, 35 men; mean age 17.5) and 111 university
students (72 women, 39 men; mean age 20) volunteered in the first experiment.13
Finnish 121
85% of the high school students and 31% of the university students were from
the capital area, the rest from different parts of Finland.14
The participants were presented with a list of thirteen noun phrases
consisting of the word tyypillinen ‘typical’ as the modifier and a morphologically
neutral human noun as the head, e.g., tyypillinen peruskoululainen ‘a typical
comprehensive school pupil’, vammainen ‘disabled/handicapped person’,
naapuri ‘neighbour’, lomailija ‘holidaymaker’.15
The participants were told to imagine the referent of the expression and
then to give a first name to the referent as quickly as possible. They were
instructed to use the same name only once and to handle each category name
individually, without comparing it to other expressions.
Female participants gave significantly more men’s than women’s names to
only two categories out of thirteen: tyypillinen suomalainen ‘typical Finn’ and
tyypillinen televisiokatselija ‘typical TV viewer’. In addition, female university
but not high school students gave significantly more men’s names to tyypillinen
lottovoittaja ‘typical lotto winner’ and tyypillinen työntekijä ‘typical worker’. The
distribution of men’s and women’s names in seven of the thirteen categories fell
short of statistical significance, e.g., tyypillinen lapsi ‘typical child’ and tyypil-
linen helsinkiläinen ‘typical citizen of Helsinki’. On the other hand, women
produced significantly more women’s than men’s names to tyypillinen potilas
‘typical patient’ and tyypillinen kristitty ‘typical Christian’.
The male participants, on the other hand, attributed significantly more
men’s than women’s names to nine of the thirteen categories. Overall, the
proportion of men’s names given for the nine categories was 86%. The male
bias was further augmented by the younger boys: Male high school students
produced significantly more men’s than women’s names for tyypillinen lapsi
‘typical child’, tyypillinen potilas ‘typical patient’ and tyypillinen kuopus ‘typical
youngest child’. None of the thirteen category names elicited significantly more
women’s than men’s names from the male participants.
The pilot for the second study was carried out in two high schools special-
ized in arts education in Helsinki. Fifty-six students (37 women, 19 men; mean
age 17.2) participated voluntarily. The participants were presented with four
short paragraphs describing ihminen ‘human being, man’ as a species, e.g., the
human being as an ‘animal capable of learning, thinking and deduction’ or the
human being as an ‘animal seeking beauty, enjoying beauty’.16
The participants were instructed to draw the ihminen featured in each
paragraph based on the images and ideas that would first come to their minds.
Secondly, they were told to give a first name to the person they had drawn
122 Mila Engelberg
(nicknames or last names were not to be used). The students were also told that
their performance would not be graded.
Male participants perceived ihminen as ‘male’ significantly more often than
as ‘female’ in three paragraphs out of four. Two paragraphs elicited exclusively
male responses. Ihminen was male in 73% of all the responses given by men.
The women’s responses showed more contextual variation than the men’s
responses. Only in one paragraph was ihminen perceived significantly more
often as ‘male’ than as ‘female’ by women. 89% of both women and men
regarded ‘human being seeking beauty’ as ‘female’. None of the four paragraphs
elicited exclusively female characters either by male or by female participants.
The results from the two studies are consonant with previous research on
covert gender and covert masculinity. The referents of both the superordinate
term ihminen and many of its morphologically gender-neutral hyponyms are
more likely to be perceived as male than as female. The interpretations are not
explicable with extralinguistic factors (cf. komissaari ‘commissar’ in example
11) or the morphological structure of the category names (cf. tavallinen
kadunmies ‘the man in the street’). Participants’ gender clearly affects interpre-
tation, however. Boys and men tend to perceive women as peripheral and
atypical members of the category ‘human’. The data indicate that the andro-
centric culture and all the relevant social information has had some effect on
girls and women as well, but they do not regard themselves as atypical humans
to the same extent as males do.
6.2 Ihminen as ‘woman’

The dictionary of Finnish dialects (Suomen murteiden sanakirja 1994; cf. also
Länsimäki 1986: 263, Rautala 1988: 33, Vilppula 1996) gives ‘woman, wife’ as
one of the meanings of ihminen ‘human being’. Ihminen denotes ‘woman’ in
addition to ‘human being’ especially in the south-western and Häme area of
Finland. It has been suggested (Vilppula 1996) that the origin of the female
connotation of ihminen in certain dialects is partly due to an asso ciation
between nainen ‘woman’ and the colloquial meaning of the verb naida ‘1. marry,
get married, 2. screw, fuck’, the first syllables of the words being identical.
According to this view, such Finnish words as naisihminen (coll., dial.) ‘female
human being’ are euphemisms for a sexualized nainen and act as a base for an
abbreviated form ihminen ‘woman’. Reasoning of this kind would seem to
require a rather conscious effort on the part of the native speaker (cf. also
Rautala 1990: 214f).
Finnish 123
Referring to a woman as ihminen (instead of nainen) is more likely to

originate from the conceptual and cultural category of ‘woman’, the sexual and
derogatory connotations of which speakers wish to avoid in this euphemistic
manner. As Mills (1995: 111) puts it: “People do not find the term man when
used for adult males potentially insulting as they do woman used for an adult
female. The term ‘woman’ itself has acquired connotations of low status and
sexuality.” The same may well apply to nainen in Finnish.
Rautala (1988: 34f) examined birth announcements in the Finnish newspa-
per Helsingin Sanomat from August 1986 till May 1987 and found that none of
the announcements of a birth of a girl (N = 127) referred to the baby as pikku
nainen ‘little woman’ or naisenalku lit. ‘beginning of a woman’, ‘little woman’,
whereas expressions such as pikku mies, pieni mies ‘little man’ or miehenalku lit.
‘beginning of a man’, ‘little man’, “were not uncommon” in announcements of
the birth of a boy (N = 154). In Rautala’s view, this reflects the unmarkedness of
man and the markedness of woman. Similarly, sentence (15a) is acceptable,
whereas (15b) evokes associations of biological and sexual characteristics:
(15) a. Niko on nyt seitsemän-vuotias urhea miehe-n-alku.
Niko be.3sg now seven-year brave man-gen-beginning
‘Niko is now a seven-year-old brave little man.’
b. Maria on nyt seitsemän-vuotias urhea
Maria be.3sg now seven-year brave
naise-n-alku.
woman-gen-beginning
‘Maria is now a seven-year-old brave little woman.’
Compounds with -nainen in old literary Finnish often have a derogatory and
sometimes a sexual connotation: jalkanainen ‘concubine’, murhanainen
‘murderwoman’, velhonainen ‘wizardwoman, sorceress, witch’ and huorinainen
‘whorewoman’ (Kiuru 1993, Engelberg 1998a). Nainen may still be more
acceptable in reference to low status groups or individuals, as in example (16):
(16) Hiv-naise-n vangitsemis-ta jatke-ttiin
hiv-woman-gen arrest-partt continue-ipf.pass
‘HIV woman still held in captivity (referring to an HIV positive Thai
woman suspected of intentionally transmitting HIV to several men)’
At the same time, it is important to note the group of compounds denoting

women where the modifier has a female reference and ihminen is the main
element: akkaihminen ‘hag’, emäntäihminen ‘matron, lady of the house,
124 Mila Engelberg
hostess’, likkaihminen ‘a slip of a girl’, piikaihminen ‘maid’, tyttöihminen ‘girl’

and vaimoihminen ‘wife’ (Suomen murteiden sanakirja 1994). Finnish has no
*miesihminen ‘male human being’, *poikaihminen ‘boy-human being’, *ukkoih-
minen ‘old man-human being’ or *isäntäihminen ‘master, head of the house-
human being’, for instance. As Koski (1983: 165f) points out, certain occupa-
tional terms have a connotation or a component [male] or [female], and they
are not normally accompanied by a modifier denoting the same gender, e.g.,
*miesteurastaja ‘male butcher’, *naiskätilö ‘female midwife’.
Perhaps expressions such as *miesteurastaja and *miesihminen would both
seem tautological. Since butchers are usually male and people=male, there is
normally no need to specify the referent’s gender. On the other hand, the
possible tautology of nouns such as *miesihminen may originate from the
male=people principle: Since men are primarily perceived as human beings and
people (not as representatives of their gender class), they are not likely to be
categorized as ‘human human beings’. Naisihminen and similar expressions are
acceptable because of the non-typicality of women as human beings and
because there is nothing gender-neutral about the concept “woman”. Ihminen
in reference to women in particular does not imply that women are perceived
as exceptionally humanlike in Finnish or in the Finnish culture.
Gender and sexism in Finnish is for the most part an unexplored area, and
further analysis is required on various levels of language and language use.
Research on generic masculinity needs to be supplemented with more experi-
ments on expressions other than occupational terms. The empirical studies
reported here on the comprehension of morphologically gender-neutral human
nouns in Finnish explored the tendency to perceive human beings as male
rather than female. Future (experimental) research should examine the other
side of the coin as well, namely the tendency to perceive men as human rather
than male. The recurrent semantic variation produced by the participants’
gender merits further study of the more unconscious and automatic level of
processing, for instance by using methods of semantic priming and reaction
time measurement.
7. Gender in Finnish proverbs
Gender and gender opposition can also be constructed and communicated

through various normative and descriptive proverbs. It is hardly surprising that
the evaluation of women from a male point of view is far more common in
Finnish 125
proverbs than the evaluation of men from women’s perspective. In his sample
of Finnish proverbs, Kuusi (1994) found that they usually evaluate women
negatively (cf. also Majapuro 1996). For example, the value of a daughter is
usually considered to be lower than that of a son:17
(17) Tytär syntyi, tyhjä syntyi, poika syntyi, kaski kaatui.
‘A daughter was born, nothing was born; a son was born, he will become
a woodland clearer.’
The “natural” power relation between women and men is spelt out in several
proverbs, e.g.:
(18) Missä akka miehen edellä, siellä reki hevosen edellä.
‘Where women go before men, there sledges go before horses.’
Majapuro’s (1997) analysis of Finnish proverbs shows that both women’s

laughter and weeping are often considered as false and calculating. Example
(19) demonstrates how a laughing woman can also be perceived as frivolous:
(19) Naisen nauru navan alta.
‘A woman’s laughter [comes] from under the belly.’
Some proverbs reflect the dominant group’s ambivalent attitude towards

women. Despite all her faults woman is still man’s life companion:
(20) Akaton mies on kuin hännätön koira.
‘A man without a woman is like a dog without a tail.’
Also, the proverbs expressing the value of a mother for her children often
convey respect or affection, e.g.:
(21) Puoliorpo isätön, koko-orpo äiditön.
‘Half-orphaned without a father, whole-orphaned without a mother.’
In Kuusi’s (1994:157) view, some proverbs dating from the 20th century may
“indicate a change in attitude” in the portrayal of women in Finnish proverbs.
Example (22) was first collected in 1844, the adaptation in (23) is from the 1950s:
(22) Ei niin huonoja housuja, jottei hametta vastaa.
‘No trousers [are] so bad that they are not the equal of a skirt.’
(23) Ei niin huonoo hametta jottei housuja vastook.
(dial.) ‘No skirt [is] so bad that it isn’t worth a pair of trousers.’
126 Mila Engelberg
8. Gendered terms of abuse
An exception to man’s greater lexical visibility as compared to woman’s is

found in derogatory sexual vocabulary. Jussila & Länsimäki (1994) have
recorded 400 words for ‘promiscuous woman’ and 120 terms for ‘promiscuous
man’. Words denoting men often have tolerant, admiring or playful connota-
tions, e.g., mela-mies ‘paddle-man’, puukko-sankari ‘[sheath-]knife-hero’.
Many animal metaphors imply strength and potency: karju ‘boar’, orhi, ori
‘stallion’. In contrast, the words for women are usually dysphemistic and more
directly derogatory: kylähuora ‘village whore’, kiertopalkinto ‘challenge cup’
and irto-liha ‘loose-flesh’. A promiscuous woman is usually conceptualized as
an object of public use.
Virtanen (1990) gathered information on the use of huora ‘whore’ in
traditional Finnish country villages based on people’s personal experience (51
women, 45 men; average age 72 years). The data were supplemented with a
small sample of university students’ responses from the 1980s. Virtanen
(1990: 158) concludes that “A whore can be a woman in general, a girl who
rejects a boy’s company, a prostitute or a woman seeing several men as well as
an unwed mother or a girl who has lost her virginity.” Virtanen also reports
how the use of huora has become more common as an expression for ‘prostitu-
te’ in the media.
Examples of pejorization of words with female reference include akka 1. ‘hag,
old woman’, 2. coll., dial. ‘wife’ and ämmä ‘hag, old woman’ which both have
denoted ‘grandmother’. Ämmä is one of the most negative words denoting
‘woman’ in contemporary Finnish (excluding sexual slurs like huora). Com-
pounds and phrases include noita-akka ‘witch-hag, witch’, juoruakka/
juoruämmä ‘gossipmonger, gossip’ and akkavalta ‘hag-power, petticoat rule’.
Piika ‘maid, hired girl’ is an old occupational title which originally meant ‘girl,
maiden’ and has now acquired a pejorative (although not a sexual) sense. The
original meaning has survived in vanhapiika ‘old maid, spinster’ and piikuus
‘maidenhood, virginity’ (Vilppula 1997).
The asymmetrical pejorization of words with female and male reference is
shown in the derivative pair naikko-nen lit. ‘woman-diminutive, trollop’ and
miekko-nen lit. ‘man-diminutive, little man’ (in reference to an adult man).
Miekkonen can be used jocularly, whereas naikkonen is clearly derogatory.
Furthermore, the male word does not comment on the referent’s sexual morals
even when used pejoratively. Insults aimed at a man sometimes compare him
to a woman, e.g., the female title neiti ‘miss, young lady’ referring to an effemi-
Finnish 127
nate man or the adjectives akkamainen ‘(old-)womanish, unmanly’ and

naismainen ‘womanish, effeminate’. A man can also be told to vetää hame
päälle ‘to put on a skirt’.
9. Female/male discourse
Women’s and men’s behaviour in interaction cannot be described on the basis

of gender alone. For example, Leiwo (1991: 158) concludes from his study on
classroom talk and gender that “Male or female dominance seems to depend on
the grade, the subject, the individual characteristics and age of the speakers and
the culture, i.e. the context.” Tainio (1996, 2001) explored shared cultural
beliefs in everyday stories and found that the relation between generations may
play a more significant role than gender relations. Other studies on interaction
have discussed, for instance, gossipping (Hakulinen 1988; Pöysä 1988), female
bonding in a conversation between young women (Haakana & Mäntynen
1993), and the relevance of gender in a TV interview (Hakulinen 1993).
10. Language reform
Gendered language has not been a public issue in Finland to the same extent as
in the United States, for example. The Union of Journalists in Finland has no
specific recommendations on avoiding sexist language in its guidelines for
good journalistic practice. When the XVIth Scandinavian Conference of
Linguistics was held in Finland in 1996, the call for papers formulated by the
Finnish organizing committee gave the following instructions: “[…] the name
of the speaker, and his university address should be given on top of the page”
[italics added].
However, sensitivity towards gendered or discriminatory language may
have increased somewhat lately and every now and then issues such as occupa-
tional titles with mies ‘man’ as the main element are discussed in the media.
Formation of morphologically neutral alternatives to gendered (occupational)
titles is in fact comparatively simple, using the suffix -jA ‘agent’, which is highly
productive in Finnish, e.g., siivota (verb) ‘to clean’ > siivoo-ja (agent noun)
‘cleaner’ or näytellä ‘to act’ > näytteli-jä ‘actor’. Similarly, a male generic title
such as marja-mies ‘berry-man, berry picker’ can be replaced with marjasta-ja
‘berry picker’ (< marjastaa ‘to pick berries’).
128 Mila Engelberg
Since Finnish lacks grammatical gender, it is not possible to use female

pro/nouns “to emphasise women’s presence in the world” the way Deborah
Cameron (1992: 126) seeks to do in “Feminism and linguistic theory”. In
Finnish, the formation of morphologically gender-neutral alternatives to
gendered expressions is a more likely strategy than the so-called feminization of
language and explicit, symmetrical gender marking in a gender-language such
as German. Finnish women themselves may regard the formation of titles with
nainen ‘woman’, for instance, as discriminatory highlighting.18
11. Conclusion
A fundamental principle of the cultural, social and linguistic gender system is

androcentricity, the perception of people as male and male as people. Andro-
centricity implicates and is based upon gender difference and gender polariza-
tion (cf. Bem 1993). After gender classes have been conceptualized as separate
and opposite categories, ‘male’ can be defined as more humanlike than ‘female’
and ‘human’ as more ‘male’ than ‘female’. Although the category of gender is
not grammaticalized in the structure of Finnish, the prototypicality of ‘male’ in
the lexicon and in the semantic system indicates that Finnish does distinguish
gender also linguistically. The more covert or implicit nature of this gender
distinction in Finnish as opposed to languages such as, e.g., French or German
does not make Finnish less gendered. Androcentricity in a genderless language
may even increase the lexical, semantic and conceptual invisibility of women.
Notes
1. Additional references on Finnish are Atkinson (1977), Branch (1987), Karlsson (1999),
Nurmi (1999), and Sovijärvi (1998).
2. The term “male” generics is preferred to “masculine” generics because Finnish lacks
grammatical gender.
3. The forms within parentheses are often used instead of hän and he, respectively.
4. The suffix -kkO is very productive in other types of derivations, e.g., poliitikko ‘politician’,
which do not specify gender. Cf. also inanimate examples such as unikko ‘poppy’.
5. The suffix -tAr has persisted in the arts, e.g., laulaja-tar ‘female singer’; kirjailija-tar ‘female
author’ is restricted to popular literature, e.g., kirjailijatar Barbara Cartland (Kyrölä 1990:10).
Finnish 129
6. Examples are taken from Finnish print and electronic media and cover the period of
1989–1998.
7. In addition, there is a spelling mistake – the phrase should read lakimiehelle tai -naiselle.
8. The collective bargaining contract for government officials (1998) consistently uses the
title virkamies and contains expressions like raskaana oleva virkamies ‘a pregnant office-man’
and directions for virkamies ‘taking maternity leave’. The corresponding contract for
municipal employees (1998), however, uses the morphologically neutral equivalent
viranhaltija ‘office-holder, civil servant’ throughout (but has retained luottamus-mies ‘trust-
man, trustee, employee representative’).
9. The acceptability of virkamies was brought up on an official level in 1986 when the law on
civil servants was being reformed and the committee for constitutional law suggested they
should look for alternatives to gender biased titles such as virkamies. Four members of the
committee, including the chair, expressed a dissenting opinion, arguing that virkamies is “in
fact gender-neutral, established and extremely apt otherwise as well” (Engelberg 1993: 45).
10. Female participants’ tendency to interpret the referents of male generics as women more
often than male participants in both studies may partly result from conscious choices.
During a debriefing of the myymäläesimies/myymälänhoitaja study in the generic male
condition, a thirteen-year-old boy said that “obviously the man part at the end of the word
highlights it [the gender bias], one thinks of a male”, to which two girls in the class replied
that they had thought of a man at first but then given a woman’s name precisely because of
that. It is not possible to conclude from the data, however, how commonly deliberate choices
of this kind are made.
11. Experimental research on male generics in a language like Finnish is complicated by the
fact that close synonyms are not always readily available (cf. generic he vs. singular they vs. he
or she in English).
12. The participants were also asked about their parents’ occupations, since this could have
influenced the responses. The proportion of housewives among both the high school and the
comprehensive school students’ parents turned out to be less than ten percent. Most of the
pupils were therefore likely to be accustomed to the idea of women in paid work.
13. One participant did not answer the age question.
14. Five high school students were from other European countries, but reported Finnish as
their native tongue.
15. Four additional noun phrases included the component [female] or [male] (tyypillinen äiti
‘a typical mother’, tyypillinen isä ‘a typical father’) or a pronounced gender bias (tyypillinen
sairaanhoitaja ‘a typical nurse’, tyypillinen taksinkuljettaja ‘a typical taxi driver’). Four animal
category names were included as fillers.
16. The paragraphs were adapted from social policy researcher Pekka Kuusi’s (1991) book
Ihmisen eloonjäämisoppi. Porvoo: WSOY.
17. The English translations in proverbs (17–18) and (20–23) are by Kuusi (1994).
18. After it became possible for Finnish women to enter voluntary military service, the media
began to use the title varusnainen ‘national servicewoman’ as opposed to the male form
varusmies ‘national serviceman’. When female draftees participating in a course at the air
130 Mila Engelberg
force technical school were informally asked about their preferences for terms of address,
they unanimously objected to any gendered references, including varusnainen ‘national
servicewoman’ and the title rouva ‘madam, ma’am, Mrs’ as in rouva kokelas ‘madam cadet’.
In fact, the women wished to be addressed in exactly the same way as the male servicemen.
For instance, they wanted to be called herra tykkimies ‘Mr artilleryman/gunner’ (Major
General Pentti Lehtimäki, personal communication 1997).
References
Atkinson, John. 1977. A Finnish grammar. 4th ed. Helsinki: Finnish Literature Society.
Bem, Sandra Lipsitz. 1993. The lenses of gender. Transforming the debate on sexual inequality.
New Haven: Yale University Press.
Branch, Michael. 1987. “Finnish.” In The world’s major languages, ed. Bernard Comrie.
London: Croom Helm, 593–617.
Braun, Friederike. 1995. “Making men out of people: The MAN principle in translating
genderless forms.” Working Papers in Language, Gender and Sexism 5: 5–38.
Braun, Friederike. 1997a. “Covert gender in Turkish.” In VIII Uluslararası Türk Dilbilimi
Konferansı Bildirileri, 7–9 Ağustos, 1996. Proceedings of the VIIIth International Confer-
ence on Turkish Linguistics, August 7- 9, 1996, eds. Kâmile İmer/N. Engin Uzun. Ankara:
Ankara Üniversitesi, 267–274.
Braun, Friederike. 1997b. “Genuslose Sprache – egalitäre Sprache? Die Behandlung der
Geschlechter in Sprachen mit und ohne Genus.” In Jahrbuch für finnisch-deutsche Literatur-
beziehungen 29. Mitteilungen aus der Deutschen Bibliothek, eds. Hans Fromm & Maria-Liisa
Nevala & Kurt Nyholm & Ingrid Schellbach-Kopra. Helsinki: Deutsche Bibliothek, 39–49.
Cameron, Deborah. 1992. Feminism and linguistic theory. 2nd ed. London: Macmillan.
Engelberg, Mila. 1993. “Maskuliini, feminiini, virkamies” [Masculine, feminine, civil
servant]. Naistutkimus-Kvinnoforskning 4: 39–50. (English abstract).
Engelberg, Mila. 1995. “Kuinka geneerisiä suomen geneeris-maskuliiniset ilmaisut ovat?”
[How generic are the generic masculine expressions in Finnish?]. Acta Universitatis
Ouluensis B Humaniora 19: 25–32.
Engelberg, Mila. 1998a. “Sukupuolistuneet ammattinimikkeet” [Gendered occupational
terms]. Virittäjä 1: 74–92. (English abstract).
Engelberg, Mila. 1998b. “Androcentricity in Finnish: Empirical findings.” Unpublished paper.
Engelberg, Milo. 2001. “Ihminen ja naisihminen: Suomen kielen piilomaskuliinisuus”
[Humans and female humans: Covert masculinity in Finnish]. Naistutkimus-
Kvinnoforskning 4: 23–26. (English abstract).
Haakana, Markku & Anne Mäntynen. 1993. “Naiset, miehet ja tuhmat jutut” [Women, men
and obscene stories]. Naistutkimus-Kvinnoforskning 4: 16–27. (English abstract).
Hakulinen, Auli. 1988. “Miehiä juoruilemassa” [Men gossiping]. In Isosuinen nainen.
Tutkielmia naisesta ja kielestä [The woman with the big mouth. Essays on woman and
language], ed. Lea Laitinen. Helsinki: Yliopistopaino, 135–156.
Finnish 131
Hakulinen, Auli. 1993. “Erään keskustelun analyysi. Lenita Airisto haastattelee Finnairin
talouspäällikköä” [Analysis of a conversation. Lenita Airisto interviews Finnair’s
economic manager]. Naistutkimus-Kvinnoforskning 4: 4–15. (English abstract).
Hakulinen, Auli & Lea Laitinen. 1993. “Kielitieteellistä naistutkimusta Suomessa” [Linguistic
women’s studies in Finland]. Naistutkimus-Kvinnoforskning 4: 60–63.
Hamilton, Mykol C. 1991. “Masculine bias in the attribution of personhood. People=Male,
Male=People.” Psychology of Women Quarterly 15: 393–402.
Husu, Liisa & Pirkko Niemelä. 1993. “Finland.” In International handbook on gender roles, ed.
Leonore Loeb Adler. Westport: Greenwood Press, 59–76.
Jussila, Raimo & Maija Länsimäki. 1994. Se siitä. Suomen kielen seksisanakirja [That’s that.
Finnish dictionary of sexual terms]. Porvoo: Söderström.
Kangas, Kea. 1997. Toisen kautta vai suoraan – henkilön määrittäminen viittauksissa [In
relation to other or directly – how are persons defined in human references?]. MA
thesis, Department of General Linguistics, Helsinki University, Finland.
Karlsson, Fred. 1974. “Sukupuoliroolien kielellisistä heijastumista” [Reflections of sex roles
in language]. Sananjalka 16: 24–33.
Karlsson, Fred. 1983. Finnish grammar. Porvoo: Söderström.
Karlsson, Fred. 1999. Finnish grammar: An essential grammar. London: Routledge.
Kiuru, Silva. 1993. “Vaimosta naiseksi vanhassa kirjasuomessa” [From wife to woman in old
literary Finnish]. Naistutkimus-Kvinnoforskning 4: 28–38. (English abstract).
Koski, Mauno. 1983. “Oliotarkoitteisten substantiivien semantiikkaa” [Semantics of nouns
with entity reference]. In Nykysuomen rakenne ja kehitys [The structure and develop-
ment of modern Finnish], eds. Auli Hakulinen & Pentti Leino. Helsinki: Suomalaisen
Kirjallisuuden Seura, 149–171.
Kuusi, Matti. 1994. Mind and form in folklore. Selected articles. Studia Fennica. Folkloristica 3.
Helsinki: Suomalaisen Kirjallisuuden Seura.
Kyrölä, Katriina. 1990. “Kaunotar elää, laamannitar on kuollut” [The beauty lives, the
lagman’s wife is dead]. Kielikello 2: 7–12.
Länsimäki, Maija. 1986. “Tässä laissa kanalla tarkoitetaan myös kukkoa” [In this law chicken
will refer to rooster as well].” In Sanalla sanoen. Pakinoita suomen kielestä [In a word.
Essays on Finnish], ed. Maija Länsimäki. Juva: Söderström, 263–266.
Lehikoinen, Leila & Silva Kiuru. 1991. Kirjasuomen kehitys [The development of literary
Finnish]. 2nd ed. Helsinki: Helsingin yliopiston suomen kielen laitos.
Leiwo, Matti. 1991. “On girls’ and boys’ discourse roles in school.” In Leikkkauspiste.
Kirjoituksia kielestä ja ihmisestä [Intersection. Writings on language and the human
being], eds. Lea Laitinen et al. Helsinki: Suomalaisen Kirjallisuuden Seura, 151–160.
Majapuro, Anne. 1996. “Zum Frauenbild in deutschen und finnischen Sprichwörtern (Darstell-
ung eines Forschungsvorhabens).” In Erikoiskielet ja käännösteoria [Special languages
and translation theory], VAKKI-symposiumi XVI. Vaasa: University of Vaasa, 147–159.
Majapuro, Anne. 1997. “Weinende Braut, lachende Frau. Geschlechtsspezifische Merkmale
des Lachens und Weinens in deutschen und finnischen Sprichwörtern.” In Laughter
down the centuries, vol. III, eds. Siegfried Jäkel & Asko Timonen & Veli-Matti Rissanen.
Turku: University of Turku, 233–248.
Mielikäinen, Aila. 1988. “Naiset puhekielen säilyttäjinä ja uudistajina” [Women as conserva-
tive users and as innovators of spoken language]. In Isosuinen nainen. Tutkielmia
</TARGET "eng">
132 Mila Engelberg
naisesta ja kielestä [The woman with the big mouth. Essays on woman and language],
ed. Lea Laitinen. Helsinki: Yliopistopaino, 92–111.
Mills, Sara. 1995. Feminist stylistics. London: Routledge.
Nuolijärvi, Pirkko. 1986. Kolmannen sukupolven kieli. Helsinkiin muuttaneiden suurten
ikäluokkien eteläpohjalaisten ja pohjoissavolaisten kielellinen sopeutuminen [The language
of the third generation. The linguistic adaptation of representatives of the large age
groups who have moved to Helsinki from southern Ostrobothnia and northern Savo].
Mänttä: Suomalaisen Kirjallisuuden Seura. (English abstract).
Nurmi, Timo. 1999. Suomen kielen sanakirja ulkomaalaisille [Finnish dictionary for foreign-
ers]. Jyväskylä: Gummerus.
Nykysuomen sanakirja [Dictionary of modern Finnish]. 1996. Lyhentämätön kansanpainos.
Neljästoista painos [14th, unabridged ed.]. Juva: Söderström.
Pöysä, Jyrki. 1988. “Kerro, kerro! Juoru-käsitteen käyttökelpoisuudesta naisten työpaik-
kakulttuurin tutkimuksessa” [Tell me! The usefulness of the concept ‘gossip’ in studying
women’s working culture]. In Isosuinen nainen. Tutkielmia naisesta ja kielestä [The
woman with the big mouth. Essays on woman and language], ed. Lea Laitinen. Helsinki:
Yliopistopaino, 115–134.
Rautala, Helena. 1988. “Jumala siunatkoon sikiötäni kuitenkin! Havaintoja suomen kielen
piilogenuksesta” [God bless my child after all! Observations on covert gender in Finnish].
In Isosuinen nainen. Tutkielmia naisesta ja kielestä [The woman with the big mouth.
Essays on woman and language], ed. Lea Laitinen. Helsinki: Yliopistopaino, 32–40.
Rautala, Helena. 1990. “Akkojen puhe ja kanojen laulu” [Old women’s talk and chickens’
singing]. In Naisen elämää [Woman’s life], ed. Kari Immonen. Helsinki: Otava, 205–261.
Sovijärvi, Sini. 1998. NTC’s compact Finnish and English dictionary. Lincolnwood, IL: NTC
Publishing Group.
Sulkala, Helena & Merja Karjalainen. 1992. Finnish. London: Routledge.
Suomen murteiden sanakirja [Dictionary of Finnish dialects]. 1994. Helsinki: Kotimaisten
kielten tutkimuskeskus.
Tainio, Liisa. 1996. “Arkikertomukset kulttuurisen tietämyksen välittäjinä” [Everyday stories
as a mediator of cultural knowledge]. Naistutkimus-Kvinnoforskning 2: 11–29. (English
abstract).
Tainio, Liisa. 2001. Puhuvan naisen paikka. Sukupuoli kulttuurisena kategoriana kielen-
käytössä [The place for a talking woman. Sex/gender as a cultural category in language
use]. Helsinki: Suomalaisen Kirjallisuuden Seura.
Väänänen, Marjatta. 1996. Suoraan eestä Suomenmaan [Straight onto Finland]. Helsinki: Otava.
Vesikansa, Jouko. 1978. Nykysuomen oppaita [Manuals of modern Finnish], vol. 2: Johdokset
[Derivatives]. Porvoo: Söderström.
Vilppula, Matti. 1996. “Murteet ja mennyt maailma” [Dialects and the bygone world].
Kielikello 2: 6–8.
Vilppula, Matti. 1997. “Piika ja menetetty maine” [Maid and the lost reputation]. Kielikello 2:
26–28.
Virtanen, Leea. 1990. “Huoraksi nimittely suomalaisessa perinteessä” [‘Whore’ in Finnish
tradition]. In Louhen sanat. Kirjoituksia kansanperinteen naisista [Louhi’s words. Essays
on women in folklore], eds. Aili Nenola & Senni Timonen. Helsinki: Suomalaisen
kirjallisuuden seura, 139–160.
<LINK "hal-n*">
<TARGET "hal" DOCINFO
AUTHOR "Kira Hall"
TITLE "“Unnatural” gender in Hindi"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
hindi
“Unnatural” gender in Hindi*
Kira Hall
University of Colorado at Boulder, USA
1. Introduction
2. Gender in the Hindi language system
2.1 Grammatical gender: Assignment and agreement
2.2 Generic masculines
2.3 Gender reversal: Terms of endearment and insult
3. Uses of the gender system by Hindi-speaking hijras
3.1 Language in hijra socialization
3.2 The exploitation of grammatical gender in everyday hijra
conversation
3.3 The use of masculine self-reference
4. Conclusions
Appendix
Notes
References
1. Introduction
Hindi (hindı̄), adopted by the Indian constitution as a national-official lan-

guage, is an Indo-Aryan language that is claimed as a mother tongue by roughly
40% of the Indian population.1 Boasting over 450 million first- and second-
language speakers worldwide, Hindi is one of the three most widely spoken
languages in the world today, along with English and Mandarin Chinese. When
speakers of Urdu are included in this estimate, a language that shares the same
basic syntax as Hindi but in certain registers draws much of its vocabulary from
Perso-Arabic sources as opposed to Sanskritic, this number is even greater.
Hindi has also been designated the official language of several states, among
them Bihar, Haryana, Himachal Pradesh, Madhya Pradesh, Rajasthan, and
Delhi; the central government has endorsed its use along with English (identified
134 Kira Hall
as the “associate official language”) in administrative functions. But these facts

obscure the complex internal politics regarding the position of Hindi in the
Indian social landscape, where several sectors of the population oppose its
ascendancy in politics and government, particularly in southern India where the
Dravidian languages Tamil and Telugu are dominant. Opposition to Hindi has
also steadily increased among Urdu speakers in response to nationalist em-
bracements of a śuddh or ‘pure’ Sanskritic Hindi by various Hindu fundamen-
talist groups, whose leaders reject Perso-Arabic influences on the language as
part of an anti-Muslim political platform. The linguistic correlate, according to
some scholars, has been an ever-increasing divergence between Hindi on the
one hand and Urdu on the other, with noncomprehensibility sometimes existing
between radical versions of each (see Shapiro & Schiffman 1983, King 1999).
Modern grammars of the many dialects now grouped together under the
label “Hindi” grew out of 18th and 19th century attempts to link Sanskrit to the
known classical and vernacular languages of Europe. Beames’ Comparative
grammar of the modern Aryan languages of India (1872–79) offers the first
taxonomy of Indo-Aryan languages and the position of Hindi therein, followed
by Kellogg’s A grammar of the Hindi language (1875) and Grierson’s highly
influential Linguistic survey of India (1903–28). Since the publication of these
early surveys, there has been constant debate regarding the appropriate taxono-
my of the languages and dialects of Indo-Aryan, a debate that points to the
impossibility of defining the difference between “language” and “dialect” in
purely linguistic terms (see Shapiro & Schiffman 1983 for an engaging summary
of this debate in the Indian context). For the state of Uttar Pradesh, where I
conducted my fieldwork, contemporary scholars generally categorize dialects of
Hindi as belonging to either Eastern Hindi or Western Hindi. The eastern
group includes the dialects of Avadhi, Bagheli, Bhojpuri, and Chattisgarhi, and
the western group Braj, Bundeli, Kanauji, and Bangru. Standard Hindi is based
on a Western Hindi dialect referred to by academics as kharı̄ » bolı̄, though only
a small percentage of Hindi speakers can be said to have spoken this variety as
a mothertongue. In addition to these dialects, there are also several local
varieties of Hindi that have developed as lingua francas or pidgins, most notably
bambaiyā hindı̄ in Bombay (Chernyshev 1971, Apte 1974) and bāzārū hindı̄ in
Calcutta (Chatterji 1931). For traditional grammars of regional dialects of
Hindi and their classification see Guru (1920), Vajpeyi (1967, 1968) and
Sharma (1958); for more contemporary descriptions of Hindi written in English
see McGregor (1972), Mohanan (1994), Shapiro (1989), and Srivastava (1969).2
Hindi 135
Although India has been an important center of sociolinguistic research

since the 1950s, fostering insightful debate on the complex relationship between
language and social identity, discussions of gender are largely absent in the
literature. This has less to do with scholarly neglect than it does with the way in
which sociolinguistics has developed on the Indian subcontinent, where caste
was established early on as the central variable of concern for the study of
linguistic variation. Following Bloch’s pioneering study “Castes et dialectes en
Tamoul” at the beginning of the 20th century (Bloch 1910), much of the early
scholarship on social stratification focused on Dravidian languages, particularly
Tamil, Tulu, and Kannada. But sociolinguistic research on Hindi also helped to
further the preoccupation with caste, when prominent linguistic anthropolo-
gists like Gumperz (1958) called our attention to caste influences on Hindi with
his important article “Dialect differences and social stratification in a North
Indian village”. Gumperz’s work stands out from many of the studies that
preceded it in that he argues for the consideration of other variables in addition
to caste, among them place of residence, religion, informal friendship contacts,
and occupation. Most critically, he brings the notion of “context” into the
sociolinguistic literature when he stresses the importance of studying patterns
of individual and group interaction.
But gender takes the backseat in Gumperz’s study, as it does in the majority
of variationist studies that followed, even when sociolinguistic scholarship on
South Asian languages broadened its focus to include variables like urbaniza-
tion, education, economic status, literacy, and age. While we might want to
blame this gap on the intellectual climate at the time – after all, the field of
language and gender did not really emerge until after the publication of Lakoff’s
“Language and Woman’s Place” in 1973 – it should be noted that Bloch himself
considered the role of gender in language use as early as 1910. He notes, for
example, that Tamil-speaking women exerted conservative influences on the
social dialects of their time, attributing this to their comparative lack of educa-
tion. In fact, Bloch points to the speech of women and men in early Sanskrit
theatre as evidence for a long history of dialect stratification in South Asia,
which often represents men of high varnas » (e.g. Brāhmana» and Kshatriya)
» as
speaking Sanskrit and men of low varnas, » servants, and women as speaking
Prakrit. Nevertheless, few articles to date have explored gender differentiation
with respect to language choice, with the notable exception of Simon’s recent
work (1993, 1996) on gendered attitudes toward the use of Hindi vs. Banarsi
Bhojpuri in Banaras.3
136 Kira Hall
Certainly the sociolinguistic literature on Hindi has involved much more

than variationist studies, especially in the past two decades where we find a
boom of research on topics ranging from code-switching and bilingualism to
language and nationalism. When we consider these other areas of inquiry,
among them the feminist analysis of sexist language (e.g. Valentine 1987) and
the pragmatic analysis of terms of address, pronoun choice, greetings, and
kinship terminology (e.g. Jain 1969, 1973; Khubchandani 1978; Mehrotra 1977,
1985a, 1985b; Vatuk 1969a, 1969b; Misra 1977), we do find a few discussions of
gender as a grammatical and as a social category. The study of address terminol-
ogy is noteworthy in this respect, particularly as its emphasis on language in
context has taken Hindi sociolinguistics in more dynamic directions. The fact
that Hindi has three second-person pronouns, for instance, whose employment
is dependent on a complex array of social considerations, disallows descriptions
of language that rely on a static conception of social role or identity. The choice
of address forms in Hindi, as the authors named above all illustrate, is not
simply a function of the social positions of addresser and addressee, predeter-
mined by the gender, marital status, and age of the conversational participants.
It is also dependent on how these social positions interact with the context of
the speech event, where the topic of discourse, intention of the speaker, degree
of intimacy and solidarity between interlocutors, emotional attitudes, and
linguistic creativity all play a role.
The study of address terminology, then, moves us squarely into the field of
discourse analysis. Yet even with this shift in research direction, there has been
little subsequent scholarship on how gender materializes in conversational
practice. Valentine’s work in the mid-1980s (1985, 1986) remains one of the few
attempts to apply contemporary interactional sociolinguistics to the Indian
context, though the bulk of her data comes from contemporary Hindi novels,
plays, and short stories. Paralleling research results for middle-class speakers of
English in the United States, Valentine finds, for example, that male speakers of
Hindi successfully initiate more conversational topics while female speakers do
more of the conversational maintenance work. But her appropriation of a two-
cultures model of gender to explain the discursive practices of communication
in mixed groups won her sharp criticism from Singh & Lele (1990), who
disagree with the way she conceptualizes power in structural-functional rather
than hierarchical terms.4
I offer this description of the speech patterns of Hindi-speaking hijras as an
example of how discourse analysis, when applied to a particular community of
practice, can reveal profound insights about the workings of gender in society.
Hindi 137
Given that the speech patterns of women and men have been so little studied
in Hindi sociolinguistics, some may disagree with my choice to focus on a
group of speakers who themselves identify as na mard na aurat ‘neither man
nor woman’. Often discussed as a “third gender” by anthropologists, most
hijras were raised as boys before taking up residence in one of India’s many
hijra communities and adopting the feminine dress, speech, and mannerisms
associated with membership. Yet the gendered liminality of these speakers is
precisely what provoked my initial interest in their language habits, since the
hijras alternate between feminine and masculine linguistic reference in ways
that reflect both local and dominant ideologies regarding the position of
women and men in north Indian society. Since 1993 I have been visiting and
researching a variety of hijra communities in northern India. Most of the data
are taken from my fieldwork in the city of Banaras, where I conducted exten-
sive interviews with hijras from four different communities and recorded their
everyday conversations. Constrained by a linguistic system which allows for
only two morphological genders, i.e., feminine and masculine, Banaras hijras
must gender themselves and fellow community members as either femi-
nine/female or masculine/male. Because nouns, verbs, adjectives, and postpos-
itions in Hindi are marked for feminine and masculine gender, with verbs
being marked in all three persons, the hijras’ attempts at alternating construc-
tions of female and male selves becomes apparent in quite basic choices of
feminine and masculine forms.
In this chapter I explore how we might go about analyzing these alterna-
tions, particularly given the traditional linguistic distinction between “gram-
matical gender” on the one hand and “natural gender” on the other. In the first
kind of system, according to conventional linguistic thought, gender is an
arbitrary grammatical category that has syntactic consequences throughout the
grammar; in the second, gender is a “natural” category that merely reflects the
“biological sex” of the referent. Feminist linguists since the 1970s (early articles
include Bodine 1975; Martyna 1980a, 1980b) have argued against the first of
these classifications, illustrating, sometimes quite convincingly, that grammati-
cal gender is not a purely arbitrary phenomenon. Others have argued against
the second of these classifications, suggesting that no classification system is
purely “natural”.5 English, for instance, while often discussed as a language
exhibiting natural gender, not only allows the usage of male generics for female
referents, it also permits the assignment of metaphorical gender to inanimates,
as when a ship, boat, or car is referred to as she. But I want to move beyond
these arguments in order to challenge the very assumption implicit in the term
138 Kira Hall
“natural gender”, i.e. that gender is a fixed phenomenon, rooted in biology and
therefore free of ideological influences. What happens to a language’s classifica-
tion system in instances when the referent’s gender can no longer be assumed
as either male or female? And what might these instances of “unnatural gender”
tell us about the relationship between gender in language and gender in society?
2. Gender in the Hindi language system
2.1 Grammatical gender: Assignment and agreement

Any discussion of the workings of gender in the Hindi language system requires
a few disclaimers. While a two-way gender system of masculine and feminine is
present in all dialects (with the exception of lingua franca or pidginized varieties
where grammatical gender is often lost altogether), the specifics of the system
materialize differently. Instantiations of grammatical gender for inanimate
nouns vary from dialect to dialect, with nouns treated as feminine in one
sometimes appearing as masculine in another, and vice versa (see Nespital
1990: 8–9 for several examples). The noun dahı̄ ‘yogurt’, for example, is often
marked as masculine in eastern dialects and feminine in western ones, a
difference most likely caused by a western reinterpretation of the final vowel -ı̄
as a feminine morphological marker. And in many of the Hindi dialects
emerging in urban areas as a result of contact with other languages, only
animate nouns referring to female individuals are treated as feminine. Original-
ly feminine nouns such as hindı̄ ‘Hindi’, nadı̄ ‘river’, and śādı̄ ‘marriage’ are
classified as masculine, a phenomenon Bhatia (1992: 174) attributes to a
“weakening of grammatical gender and the preference for the natural gender”.
In other words, nouns in these dialects are normally not feminine unless they
specifically reference female persons.
Moreover, while speakers of Hindi in the kharı̄ » bolı̄ area6 most closely
produce the kind of gender agreement outlined in instructional grammars,
speakers of certain eastern varieties often perceive this same gender agreement
as stylistically marked when employed across syntactic distance. Because Hindi
is a kind of “communication amalgam” (Khubchandani 1991: 273), speakers
generally exhibit neutral attitudes toward variations in speech, grammatical
gender notwithstanding. But the gender variability described above leads Simon
(ms.) to make the interesting claim that the employment or non-employment
of standard gender agreement sometimes serves as a register marker, indexing
Hindi 139
the speaker’s gender, linguistic proficiency, education level, and/or insid-

er/outsider status. Thus in the western Hindi area, according to Simon, elite
speakers of standard Hindi will often perceive a speaker who uses non-standard
gender agreement as purbı̄ (a person from eastern Uttar Pradesh) or dehātı̄ (a
villager, a rural illiterate). But in Banaras, where the local dialect of Banarsi
Bhojpuri is spoken along with standard Hindi, it is the use of standard gender
agreement that is sometimes deemed suspect. Simon offers evidence that this is
particularly the case among uneducated female speakers of the local dialect,
who frequently perceive speakers who maintain the strict gender agreement of
standard Hindi as either foreign or over-educated.
Simon’s observations regarding Banarsi Bhojpuri cannot easily be applied
to the Banaras hijra community, however, since its members rarely come from
the city. Hijras are notorious travelers, moving from state to state, city to city,
and community to community for years before they settle in any one place.
Indeed, their life narratives are constructed around movement: Because so
many of them were forced out of their homes at an early age for exhibiting
behavior deemed to be “hijra-like” or effeminate, their stories reveal an ongoing
state of homelessness and displacement. Those hijras that eventually come to
settle in Banaras, then, are rarely native speakers of Banarsi Bhojpuri, and its
employment in the community is usually quite marked. What we find instead
is a kind of lingua franca loosely based on standard Hindi, which facilitates
communication among speakers from a variety of regional and linguistic
» bolı̄ in Banaras,7 most
backgrounds. The hijras’ language, which they call hijrā
closely parallels those varieties of Hindi designated by some Hindi speakers as
» or milijhulı̄ (see, for example, Sachdeva 1982), terms which translate
khicrı̄
roughly as ‘mixed’. Such varieties, among them the Panjabi-ized Hindustani
spoken around Delhi, usually develop on the borders of language or dialect
areas and reflect features from the divergent bordering varieties. In this context,
» a term coined by the
Hijra Boli might best be thought of as a kind of sadhukkarı̄,
medieval mystical poet Kabir for the ‘mixed’ language of travelers (literally, the
language of sādhus or religious mendicants).8 This term is especially fitting for
the hijras’ language, given the hijras’ traditional religious role at birth and
wedding celebrations, where their blessing is thought to secure a long and
fruitful lineage of sons for the recipient.
But where Hijra Boli differs from the lingua francas or pidgins reported for,
say, the streets of Bombay and Calcutta is in the use of grammatical gender. In
contrast to those varieties, which exhibit a loss of grammatical gender, the
variety of Hindi adopted by the hijras tends to overemphasize gender, using
140 Kira Hall
masculine and feminine gender in places where it would normally not appear
in kharı̄» bolı̄, or treating nouns that are masculine in standard Hindi as feminine
and vice versa. To give one illuminating example: The word hijrā » is grammati-
cally masculine in standard Hindi, but the hijras frequently treat the noun as
feminine through verbal agreement when it acts as the subject of a sentence. As
I argue here, this usage reflects a kind of gender overcompensation or even
hypercorrection. Upon entering the community, Banaras hijras work to
distance themselves from masculine representations, with many of them even
choosing to undergo a ritualized penectomy and castration operation. The fact
that the term hijrā » is grammatically masculine sometimes gets in the way of this
communal distancing, so hijras will mark the noun as feminine as part and
parcel of “doing gender”. The two-gender system exhibited in standard Hindi,
then, is quite relevant to a discussion of the language practices of this commu-
nity, since grammatical gender is most often overemphasized, not underempha-
sized, in the hijras’ constructions of a more feminine self.
The alternation between feminine and masculine reference in standard
Hindi is quite easy to discern linguistically, since many nouns, verbs, post-
positions, and adjectival modifiers inflect for gender. Nominals, for instance,
exhibit both a two-way gender system of masculine and feminine as well as a
two-way number system of singular and plural. While the gender of animate
nouns to a certain extent corresponds to the referent’s gender (i.e. exhibiting
“natural gender”), gender designation in inanimate nouns is comparatively
arbitrary (though see Valentine 1987 for a discussion of the sexist basis of some
of these designations, and Börner-Westphal 1989 for a discussion of lexical gaps
in the gender system). Hindi nominal forms are classified as either direct
(nominative) or oblique (non-nominative), with the latter normally signaled by
the presence of a postposition. For the majority of nouns in the direct case, the
-a ending signals masculine singular, -e masculine plural, -ı̄ feminine singular,
and -iyÖa feminine plural; in the oblique case these endings become -e, -õ, -ı̄,
and -iyõ, respectively. Not all adjectival modifiers exhibit inflection, but those
that do agree with their head noun in gender, number, and case. Masculine
forms of inflecting adjectives end in -ā in the singular direct and -e in the plural
direct, singular oblique, and plural oblique cases; the feminine forms always end
in -ı̄, whether singular or plural, direct or oblique. Examples illustrating gender
agreement in nominals are given in (1) and (2):
Hindi 141
(1) Masculine agreement

a. acchā »
larkā
good.nom.masc.sg boy.nom.masc.sg
‘good boy’
b. acche »
larke
good.nom.masc.pl boy.nom.masc.pl
‘good boys’
c. acche »
larke ko
good.obl.masc.sg boy.obl.masc.sg to
‘to the good boy’
d. acche »
larkõ ko
good.obl.masc.pl boy.obl.masc.pl to
‘to the good boys’
(2) Feminine agreement
a. acchı̄ »
larkı̄
good.nom.fem.sg girl.nom.fem.sg
‘good girl’
b. acchı̄ » a
larkiyÖ
good.nom.fem.pl girl.nom.fem.pl
‘good girls’
c. acchı̄ »
larkı̄ ko
good.obl.fem.sg girl.obl.fem.sg to
‘to the good girl’
d. acchı̄ »
larkiyõ ko
good.obl.fem.pl girl.obl.fem.pl to
‘to the good girls’
The genitive postposition also agrees with the gender of the head noun,
appearing as kā when modifying a singular masculine noun, ke when modifying
a plural masculine noun, and kı̄ when modifying a singular or plural feminine
noun. For concentrated discussions of grammatical gender in Hindi nominals,
see Pathak (1976) and Gosvāmı̄ (1979).
Verbals in standard Hindi also show gender agreement, agreeing with the
subject in gender, number, and person if it is in the nominative case. If the
subject is in an oblique case – i.e. ergative, dative, instrumental, locative, or
genitive – the verb agrees with the object when it is nominative.9 In general, the
appearance of one of the vowels -ā, -e, -ı̄, or -Öı on the verb signals number and
gender, with -ā used for masculine singular, -e for masculine plural, -ı̄ for
feminine singular, and -Öı for feminine plural. For example, the intransitive verb
142 Kira Hall
honā ‘to be’ is realized as thā with masculine singular controllers, the with
masculine plural controllers, thı̄ with feminine singular controllers, and thÖı with
feminine plural controllers, as illustrated in Table 1.
Table 1.Past tense forms of honā ‘to be’

Masculine Feminine English Translation
sg 1 maı̃ thā maı̃ thı̄ I was

2 tū thā tū thı̄ you (intimate) were
3 vah thā vah thı̄ he was/she was
pl 1 ham the ham thÖı we were
2 tum the tum thÖı you (familiar) were
3 ve/ap the ve/ap thÖı they/you (polite) were
I should add that many speakers of standard Hindi do not employ the
feminine plural ending -Öı with any regularity, particularly when the subject of
the sentence is in the oblique case and the verb therefore agrees with the object,
as in the sentence reproduced in example (3):
(3) a. Standard Hindi
āp-ne kitnı̄ kitābẽ
you.pol-erg how.many.nom.fem book.nom.fem.pl
» ı?
parh-Ö
read.pf-fem.pl
‘How many books did you read?’
b. Colloquial Hindi
āp-ne kitnı̄ kitāb
you.pol-erg how.many.nom.fem book.nom.fem.sg
»
parh-ı̄/pa » ı?
rh-Ö
read.pf-fem.sg/pl
‘How many books did you read?’
Here, where the second person subject is in the ergative case, even standard
Hindi speakers will frequently say ‘book’ (kitāb) instead of ‘books’ (kitābẽ) and
employ the singular feminine verbal ending in agreement, as in (3b). Again,
examples like this point to the fact that when it comes to conversational
practice, there is no singular way to explain “gender in Hindi”.
I can only hint at the complexity of the Hindi verbal system, since auxiliaries
and modals combine in various ways with either the verb root or its inflected
Hindi 143
forms to yield numerous distinctions of tense, aspect, mood, and voice. In the
imperfective, continuous, and perfective verb forms, aspect is indicated through
the addition of explicit markers of various kinds to the stem while tense is
indicated through the presence of one of the basic forms of honā ‘to be’ (i.e.
present, past, presumptive, subjunctive). Here, too, the appearance of one of the
vowels -ā, -e, -ı̄, or -Öı will signal the gender and number of the NP with which the
verb agrees, and each element in the complex that is not in the root form will
reflect these agreement features. First-person feminine and masculine agreement
for several tenses of the verb jānā ‘to go’ are illustrated in Table 2.
Table 2.Selected examples of first person verbal marking with jānā ‘to go’
Verb Tense 1st Person Masculine 1st Person Feminine English Translation
Future maı̃ jāÖugā maı̃ jāÖugı̄ I will go

Past maı̃ gayā maı̃ gayı̄ I went (definite)
General Present maı̃ jātā hÖu maı̃ jātı̄ hÖu I go
Imperfective Past maı̃ jātā thā maı̃ jātı̄ thı̄ I went (indefinite)
Continuous Present maı̃ jā rahā hÖu maı̃ jā rahı̄ hÖu I am going
Continuous Past maı̃ jā rahā thā maı̃ jā rahı̄ thı̄ I was going
Perfective Present maı̃ gayā hÖu maı̃ gayı̄ hÖu I have gone
Perfective Past maı̃ gayā thā maı̃ gayı̄ thı̄ I had gone
The import of this kind of verbal morphology for the hijras, of course, is
that even when pronouncing simple first-person statements like ‘I am going’,
hijras must gender themselves as either feminine or masculine.

The hijras’ alternating uses of feminine and masculine morphological forms for
the same referent appears in very limited contexts in non-hijra Hindi-speaking
communities. With perhaps the exception of second language learners unfamil-
iar with the gender system, Hindi speakers rarely, if ever, betray their “natural”
gender when referring to themselves. Likewise, they generally respect the gender
of the referent or addressee in question – except, of course, when the generic
masculine is used as inclusive of both female and male persons, as in the nouns
reproduced in (4):
144 Kira Hall
(4) dost ‘friend (male or female)’

sāthı̄ ‘companion (male or female)’
mitra ‘friend, ally (male or female)’
yātrı̄ ‘traveler (male or female)’
»
ghursavār ‘rider (male or female)’
Feminine counterparts to many of the masculine nouns in (4) exist, such as

sahelı̄ for ‘female friend’, but these terms are never used generically; rather, their
usage expresses specific reference to a female individual. Most females will use
sahelı̄ for their female friends, for instance, and will rarely call a female friend
dost. The same is true with respect to verbal agreement with a pronoun like koı̄
‘someone’. As the two sentences in (5) illustrate, masculine verbal agreement
with koı̄ could point to either a male or female subject, but feminine verbal
agreement will always denote a female subject.
(5) a. koı̄ āyā hai
someone come.pf.masc.sg be.pres
‘Someone (male or female) has come.’
b. koı̄ ayı̄ hai
someone come.pf.fem.sg be.pres
‘Someone (female) has come.’
And so it is that we find the generic masculine in proverbs such as those

reproduced in (6):
(6) a. sab ādmı̄ barābar haı̃
all man.masc.pl equal be.pres
‘All men are equal (i.e. all people are equal).’
b. har manusya » kā
every man.obl.masc.sg gen.masc.sg
dharm hai
duty.nom.masc be.pres
‘Every man has his duty (i.e. every person has his/her duty).’
c. jo boyegā so ka»tegā
who sow.fut.masc.sg he reap.fut.masc.sg
‘He who sows will reap.’
There is a limited set of feminine epicene terms in standard Hindi, among them
the term savārı̄ (f) ‘passenger’ which is used for both male and female referents,
but this is comparatively rare, cf. (7):
Hindi 145
(7) savārı̄ āyı̄

passenger.fem.sg come.pf.fem.sg
‘The passenger came.’
As with many of the languages discussed in Gender across Languages, such

examples point to an asymmetry in the grammatical gender system of Hindi, a
point Valentine (1987) made over a decade ago from a feminist perspective.
This asymmetry is also reflected in the verbal agreement in sentences with a
complex subject that includes both a masculine and a feminine animate noun;
in such cases, the masculine form is always used. When the complex subject
involves inanimate objects of different genders, however, the verb will normally
agree with the noun nearest to it in word order.
2.3 Gender reversal: Terms of endearment and insult

Outside of these uses of the generic masculine, however, speakers generally
match morphological gender with referential gender. But there are a few
notable exceptions. One of these is the occasional use of the masculine term
be»tā (m) ‘boy’ instead of the feminine be»tı̄ (f) ‘girl’ in direct address to a
younger woman or daughter. When used in this way, particularly by parents to
their children, the term becomes a term of endearment and best translates as
‘dear’. One might argue that this gender reversal works as endearment because
of the traditional value given to sons, as opposed to daughters, in Indian
culture. When masculine terms are used for female persons, then, they tend to
elevate the status of the referent.10 So when women use the term bhāı̄ (m)
‘brother’ for one another, as often happens between intimates, it signals equality
and informality (cf. also Tobin, vol. I for gender switch in Hebrew).11 A parallel
might be drawn here with the honorific masculine term sāhab (m) ‘sir’, which
is frequently used in reference to women, as in memsāhab (m) ‘lady’, dāk»tar
sāhab (m) ‘doctor’, and profesar sāhab (m) ‘professor’ (see Valentine 1987 for
further discussion). These masculine address terms are highly respectful when
used for women, underscoring the recipient’s social status. When Indian
journalists referred to Mrs. Gandhi as “the only man in the cabinet”, they were
highlighting her ability to run the affairs of the state (and the inability of male
cabinet members to do so).12
But male persons are rarely addressed or referenced in the feminine gender.
There are a number of insult terms which “imply” effeminacy and act as an
insult to the recipient’s masculinity, among them the term hijrā » when used by
146 Kira Hall
non-hijra men. The term, which literally means ‘impotent’, is frequently

employed either in joking or anger to indicate the ineffectiveness of the referent
in question. Other masculine terms of insult that connote effeminacy are
included in (8). The hijras I knew in Banaras all, with much grief, reported
having been called several of these names as young children (see Hall 1997).
(8) Selected masculine insult terms used for males
»
hijrā ‘eunuch, impotent’
nacaniyÖa ‘little dancer’
bhosrı̄» vālā ‘vagina-owner’
chakkā ‘a set of six, effeminate man’
chah nambar ‘number six, effeminate man’
»
gāndū ‘passive partner in sodomy’
jankhā ‘effeminate man’
These terms are all grammatically masculine, and this again points to an
asymmetry in the Hindi gender system. While the use of masculine terms for
female persons does occur and tends to carry positive implications, the use of
feminine terms for male persons is virtually non-existent,13 and when feminini-
ty (or effeminacy) is implied, as in the terms listed in (8) above, the connota-
tions are extremely negative.
It is much less common for a male speaker to make use of the first-person
feminine or a female speaker to employ the first-person masculine. There are
infrequent accounts of young girls who, as a result of spending their formative
years playing in predominantly male environments, speak in the masculine
until taught to do otherwise in the Indian school system. But it is much rarer to
find a boy speaking as a girl, particularly when there is such stigma attached to
doing so. This might explain why the Hindi speakers in Banaras who reviewed
my transcripts of hijra conversations were shocked at the hijras’ use of feminine
forms for themselves and other community members. Although the hijras
produce an exaggerated feminine ideal in their mannerisms and dress, most
non-hijras nevertheless consider them male and normally refer to them in the
masculine gender (see Hall & O’Donovan 1996). The idea that hijras use
feminine self-reference comes as a complete surprise to many Hindi speakers,
who see the claiming of feminine morphology by speakers they identify as male
(albeit “inadequately” male) as highly abnormal.14 As I argue in the subsequent
two sections, the hijras’ varied uses of feminine and masculine first, second, and
third person verbal forms reflects a unique dual-gender position in a society
that views them as neither fully female nor fully male.
Hindi 147
3. Uses of the gender system by Hindi-speaking hijras
When a hijra joins a Banaras hijra community, she quickly learns a variety of
ways to distance herself from the masculine semiotics in which she was raised:
She begins to wear traditional feminine dress (e.g. saris, jewelry, make-up),
adopt hijra gestures (such as their distinctive flat-handed clap), sing and dance
like a hijra, and most critically for this article, speak in a more feminine
manner. Language is a critical component of this second gender socialization
process, so much so that the hijras readily distinguish between what they refer
to as mardānā bolı̄ ‘masculine speech’ on the one hand and zanānā bolı̄ ‘femi-
nine speech’ on the other, with the latter variety accepted as the preferred way
of speaking. The precise definition of these two terms varies among the hijra
communities I worked with in Banaras, but members generally agree that
mardānā bolı̄ involves direct speech, “stronger” curses, and masculine first
person verb forms, while zanānā bolı̄ entails indirect speech (and is therefore
deemed more polite), “weaker” curses, and feminine first-person verb forms.
Whether or not these qualities play out in the ordinary speech of non-hijra men
and women has not yet been studied, but the hijras are no doubt identifying
stereotypical perceptions of masculine and feminine behaviors in dominant
north Indian culture (which, incidentally, are not altogether different from
those identified for many middle-class European-American communities). But
the hijras accept these distinctions as fact, and use them to instruct each other
on how to build a “less masculine” gender presentation.
3.1 Language in hijra socialization

The hijras’ discussions of their socialization into the community is extensive in
my data, but I offer typical excerpts from these discussions here so as to give
some indication of the importance of language to this process. The first is taken
from Rupa,15 a Banaras hijra who came to the community at the comparatively
late age of 18 and found the change from mardānā to zanānā bolı̄ particularly
difficult. Raised in a prestigious Brahman family, Rupa spent all of her boyhood
conforming to masculine roles and representations. But she never felt comfort-
able with her gender and ultimately decided to join the hijra community,
undergo a castration operation, and take on the religious role of pūjārı̄ (Hindu
priest). As she describes in the following passage (9), the acquisition of “femi-
nine speech” was an especially long and laborious process. (The transcription
symbols used here are identified in the Appendix.)
148 Kira Hall
(9) “Changing that takes time” Rupa 1993
Rupa: ghar mẽ, to – R: At home –

mardānā rahatem them, they werem livingm in a mardānā way,
m
to mardānā bolı̄ bolte-bolte haı̃. so they’re always speakingm mardānā bolı̄.
jab hijre » ko jānā partā
» hai When a hijra has to leave,
to parivartan karnā partā » hai. … then a change has to be made. …
vahı̄ to bolā, na bet» ā? – That’s exactly what I told you, right dear? –
jab ghar se calem, – When I leftm home, –
jab ghar se āyem, when I camem from home,
to ghar kı̄ bolı̄ mardānā to mẽ, the speech at home was mardānā so I,
to mardānā bolı̄ bolām. (3.0) so I spokem mardānā bolı̄.
Kira: » kı̄ tarah se āpas mẽ vyavhār karte the? K: You behaved like boys with each other?
larkõ
Rupa: hÖa. bhaı̄yām ko “bhaı̄yām” bol rahem haı̃, – R: Yes, they’re callingm brothersm “bro-
cācām ko “cācām” bol rahem haı̃, – thersm”.
m
aise bol rahe haı̃. (2.0) They’re callingm their unclem “unclem”.
to usko parivartan karne mẽ They’re speakingm like that.
to t» āı̄m lagta hı̄ hai. (2.0) So changing that
to usko parivartan karne mẽ just takes time.
»
tāı̄m lagta hai. – Changing that
bolte-bolte bolte-bolte, takes time.
ādat ho gayı̄ (1.5) But gradually after speaking continously,
sāt-chah mahı̄ne mẽ. it became a habit
in about six or seven months.
Here, Rupa is in part illustrating the same-sex nature of socializing in her home
village, where men associate with men (e.g. brothers and uncles) and women
with women. Her entry into the hijra community involved a shift in play-group,
so to speak, as she suddenly found her primary companions identifying as
feminine and shunning masculine self-reference altogether. Rupa’s transition
from masculine to feminine speech, then, was a highly conscious process, one
that required several months of practice. In Rupa’s own words, it was only after
bolte-bolte bolte-bolte ‘speaking and speaking, speaking and speaking’ before it
ādat ho gayı̄ ‘became a habit’.
What is even more interesting in the hijras’ discussions of this socialization
process, however, is their frequent conflation of zanānā bolı̄ with intimacy and
solidarity and mardānā bolı̄ with social distance. In their discussions with me,
the hijras often offered examples of how they talk with one another, listing a
host of feminine-marked first and second person verb forms. But in many of
these discussions, the hijras also point to the use of the intimate second person
pronoun tū as an example of zanānā bolı̄, a juxtaposition that suggests an
association of feminine speech with intimacy. In excerpt (10), for example,
Hindi 149
Megha lists both kinds of phrases as examples of what she calls auratõ kı̄ bolı̄,
literally the ‘speech of women’:
(10) “We always speak women’s speech” Megha 1993
Megha: hÖa, hameśā auratõ kı̄ bolı̄ boltı̄f haı̃ M: Yes, we always speakf women’s speech.
kabhı̄ bhı̄ ādmı̄ ke jaisā nahÖı boltı̄f haı̃, – jaise, We never ever speakf like a man.
“maı̃ jā rahı̄f hÖu jı̄”, It’s like, “I’m goingf sir/mam”,
f
“jā rahı̄ bahan”, “Sister is goingf”,
intimate
“tū khā le”, “You (intimate) eat!”
“tūintimate pakā le”, “You (intimate) cook!”
Ö
“maı̃ abhı̄ ā rahı̄f hu”. “I’m comingf now”.
Unlike Rupa, who lives in a house with an Indian family and dresses like a man
in her “off-hours”, Megha lives with a community of hijras and sees any
admission of masculinity as a threat to her already marginalized status as a hijra
(hence the strength of her assertion “we always speak women’s speech; we never
speak like men”). The very fact that Rupa in excerpt (9) refers to both herself
and fellow community members in the masculine undermines Megha’s claim,
and points to divergent embracements of feminine self-reference by community
members. Although speaking zanānā bolı̄ is clearly the norm for hijras living in
the four Banaras communities I researched (even Rupa uses feminine self-
reference when discussing her activities with the community), members do not
always use feminine forms for themselves and others, as Megha claims here.
When a speaker wishes to express social distance between herself and another
community member, whether it be out of respect (as for her guru) or disgust
(as for an estranged hijra), she will often refer to that member in the masculine.
Moreover, the hijras will sometimes use the first-person masculine for reasons
of emphasis – for example, when angry or upset, when stressing a particular
point, or when referring to their pre-hijra selves.
This again reminds us of the care we must take with folk-linguistic observa-
tions, as perceived speech patterns rarely match actual language use. But the
hijras’ perceptions of their language habits are nevertheless worthy of study, for
the ideologies of gender uncovered therein influence their discursive interac-
tions in important ways. In excerpt (11), for instance, we find Sulekha errone-
ously claiming that she speaks like a woman when speaking with a woman and
like a man when speaking with a man:
150 Kira Hall
(11) “I do the same speech of the person I meet” Sulekha 1993
Sulekha: mujh ko koı̄ bāt nahÖı rahatā hai, S: It’s just not a big deal to me.
Ö –
maı̃ aurat jaisı̄ boltı̄f hu, [Normally] I speakf like a woman,
Ö –
ādmı̄ se ādmı̄ jaisā bāt kartı̄f hu, [but] with a man I speakf like a man.
f
Ö …
jo jaisā miltā hai us se bāt kartı̄ hu, I dof the same speech of the person I meet. …
jaise ab ham- hai na? - For example, take my case, okay?
ab- ab- auratõ mẽ haı̃, (0.5) If I’m socializing with women
to – aurat ā gayı̄ to aurat vālā hı̄ boluÖ gı̄f, and a woman comes by I’ll just speakf like a
woman.
“dı̄dı̄ bahan” kahÖugı̄f. – I’ll sayf, “Dı̄dı̄! Bahan!”
ādmı̄ ā jātā hai to If a man comes by [I’ll say]
((softly)) “kyā khāte haı̃polite. (1.0) ((softly)), “What are you (polite) eating?
kyā bāt hai āpkopolite. (1.0) What’s the matter, sir (polite)?
kyā kām hai”. What brings you here?”
By the end of the passage it becomes clear that what Sulekha means is that she
makes her speech correspond to the social distance she perceives as existing
between herself and her addressee. Women, for Sulekha, appear to represent
intimacy and informality, and so she quotes herself as using familiar terms of
address like dı̄dı̄ and bahan when speaking with them, both of which translate
as ‘sister’ (the first suggesting respect and intimacy, the second equality and
informality). In contrast, men appear to represent distance and formality, hence
Sulekha’s repeated uses of the polite second person plural āp when reproducing
her conversations with them. This short passage, even though it reveals little (if
anything) about Sulekha’s actual language use, does suggest that Sulekha, like
Megha, associates feminine speech with intimacy and masculine speech with
social distance.
It is this association that governs the choice of feminine or masculine
forms among community members. Because hijras are considered “neither
men nor women”, they have at their disposal an added resource for marking
social relations – that of grammatical gender. Indeed, their use of the gender
system in many respects parallels the use of second person pronouns in the
general population, where the choice of intimate tū, familiar tum, or polite āp
indicates the respective social positions of addresser and addressee, as well as
a host of other social attitudes and dimensions dependent on the context of the
speech event. The pronoun tū, for example, is used in cases of heightened
intimacy, e.g. to address a god, to address a close friend of equal status, to call
a small child, to express anger or disgust. On the opposite end of the relational
scale, the pronoun āp is used in situations involving any degree of formality,
particularly in situations of social inequality when the speaker wishes to signal
Hindi 151
respect for the addressee – e.g. for a guru, an elder, a parent, an employer.
Between these two extremes we find the pronoun tum, a form most often used
in informal situations by friends and colleagues but also in situations of social
inequality; for example, when an individual addresses someone of lesser status.
Gender also plays an important role in pronominal choice: A classic example
comes from the traditional Hindu family, where the husband will use the
intimate tū or tum when addressing his wife while she will use the formal āp
when addressing him. As this brief discussion suggests, the social rules govern-
ing the varying employments of these three pronouns are terribly complex, and
speakers create social relations with pronominal choice as much as they affirm
them. The same is true with the hijras’ choice of feminine or masculine
reference for members of their own community. Equipped with an extra
linguistic resource, the hijras have developed their own system for marking
social relations, one that has gender at its center.
3.2 The exploitation of grammatical gender in everyday hijra

conversation
In the next few pages, I offer selected examples of how the hijras exploit the
grammatical gender system in their everyday conversations. Since feminine
reference is expected within the community (the hijras even take on female
names when initiated), I focus on those conversational excerpts which diverge
from this expectation – that is, when the hijras use masculine reference for
themselves and for one another. The hijras regularly use feminine reference to
express solidarity with the referent in question, in the same manner that the
familiar pronoun tum is used among good friends. Fellow celā’s ‘disciples’
quickly learn to use feminine reference for one another after joining the
community; they will also use feminine reference for a superior or inferior
when wishing to show affection. But when status is a point of emphasis,
masculine reference is favored. The hijra community relies upon elaborate
familial structures which delegate various feminine roles to different members
of the group, among them dādı̄ ‘paternal grandmother’, nānı̄ ‘maternal grand-
mother’, mausı̄ ‘mother’s sister’, cācı̄ ‘uncle’s wife’, and bahin ‘sister’. But these
designations are also extremely hierarchical, with elders enjoying the respect of
newcomers and both young and old members deferring to the community
guru. The guru-chela relationship is fundamental to the hijra kinship system,
and in any given group one will find a guru surrounded by a hierarchy of chelas,
grand-chelas, great grand-chelas, and great great grand-chelas. While chelas will
152 Kira Hall
normally use feminine forms when addressing a superior directly, they will
often use masculine reference when talking about her in the third person so as
to mark their respect for her position. Conversely, hijras will frequently refer to
newer chelas in the masculine, first to differentiate them as inferiors and second
to indicate that they are still in the learning stages of hijrahood. This use of
masculine reference, then, could be said to parallel the use of the intimate
second person pronoun tū in the larger population, which can be used to
address both a god and a servant. What we have here, then, is a neat linguistic
pattern whereby feminine forms are used for hijras considered social equals and
masculine forms are used for hijras of either higher or lower status.
Because the hijras associate the masculine with both hierarchy and social
distance, they also employ it to express their dissatisfaction with other hijras.
This kind of masculine reference can readily be seen in their uses of the hijra
naming system. When a new member enters the hijra community, she is given a
woman’s name to replace the name of her former, more male self. The hijras are
discouraged from referring to each other with these remnants of their previous
lives, yet tellingly, they often employ them in disputes. If a hijra is in a fierce
argument with another member of her community, one of the most incisive
insults she can give is to question her addressee’s femininity by using her male
name. Likewise, in example (12), we see Sulekha insulting Muslim hijras by
referring to them in the third-person masculine. Although Muslims and Hindu
hijras often live together harmoniously in the same communities – an arrange-
ment rarely found in mainstream Banaras where the tension between Muslims
and Hindus is quite pervasive – Sulekha, raised Hindu, feels somewhat threat-
ened by Muslim hijras, as they hold powerful positions within the Banaras hijra
network, and indeed, throughout all of northern India. The distance Sulekha
feels towards Muslim hijras is reflected in her use of third person masculine verb
forms when Muslim hijras act as subjects, as in the following exchange between
her and my research assistant Vinita:
Hindi 153
(12) “I’d be a small mouth with big talk” Sulekha 1993
»
Sulekha: bazardı̄hā mẽ jo channū hai, S: That Channu who lives in Bazardı̄hā»
to vah bhı̄ ādmı̄m hai. is a manm.
» to hai nahÖı . …
hijrā He’s not a hijra. …
vo buzurg hai. He’s very old.
vah sab se mālikm vahı̄ hai. (1.0) He’s the chief masterm over there,
sab se mālikm vahı̄ hai. the chief masterm over there.
»
bazardı̄hā kām. - »
Ofm Bazardı̄hā.
ye vo channū iske sab ādmı̄m haı̃, All of the ones under Channu are menm,
sab ātemhaı̃ jātem haı̃. all of them who comem and gom over there.
kurtā lungı̄ pahan letem haı̃, They wearm kurtās and lungı̄s,
nācne samay sārı̄» pahan letem haı̃, - but when they dance they wearm sarees.
m
sabhı̄ jānte haı̃, (2.0) Everybody knowsm it
maı̃ hamko kahne se kyā? so what’s the use of my saying so?
Vinita: lekin vo sab āpreśan karāye hue haÖı ? V: But haven’t they all had operations?
Sulekha: nahÖı S: No.
Vinita: kuch nahÖı <hai?> V: Nothing at all?
Sulekha: <nahÖı .> S: No.
Vinita: tabhı̄ aisı̄ <hai?> V: So they’re just that way?
Sulekha: <hÖa.> S: Yes.
Vinita: o:::h. (1.0) acchā? V: O:::h. Really?
Sulekha: usko- unko maı̃ kaise kahÖu? S: How can I say anything about them?
Ö f to merā bāt kā»t dẽgem. …
usko kahugı̄ If I’d sayf anything, they’d just contradictm me
Ö f (0.5) to
maı̃ kah dugı̄ anyway. If I’d givef anything away then
((softly)) maı̃ cho»te mũh barı̄ » bāt, ((softly)) I’[d be] a small mouth with big talk.
hamko isı̄ mẽ rahnā hai. (0.5) I have to live in this [community], after all.
sab māregām pı̄t» egām bāl kā»t degām. They’d all hitm me, beatm me up, cutm my hair.
Sulekha’s use of the third-person masculine to describe the 78-year-old Channu

stands in stark contrast to her repeated use of feminine forms for herself and
fellow community members in other conversations. But Sulekha appears to
view Muslims as below her on the social hierarchy, evidenced in her insistence
throughout her interviews with us that Hindu hijras existed long before Muslim
hijras, and moreover, that it is only hijras from low caste backgrounds who
convert to Islam and eat meat. Displeased with her own “smallness” relative to
these Muslim hijras, Sulekha refuses to grant the entire community any
acknowledgement of femininity, whether it be linguistic or anatomical.16
A comparable instance of such distancing can be found in Megha’s refer-
ences to Sulekha. After a fairly serious argument with Megha, Sulekha left
Megha’s community in Banaras and went to live with a male partner in a
neighboring village outside the city. In a manner consistent with her claims in
(10), Megha almost always uses feminine forms when referring to other hijras;
154 Kira Hall
yet when she refers to Sulekha, who apparently insulted her authority as mālkin
(f) ‘chief ’ of her community, Megha uses the masculine. Two examples of this
employment are reproduced in excerpt (13):
(13) “Now he left” Megha 1993
Megha: bacpan se yahÖı kām hai, – M: He belongedm to this household since

childhood,
ab jākar [place name] mẽ rah rahām hai, –
now he left and is livingm in [place name].
merā jajmānı̄ hai,
I had clients there,
Ö
to maı̃ un logõ ko de detı̄f hu.
but I transferredf those people to him.
Through the use of masculine postpositions like kā (m) ‘of (m)’ and masculine
verb forms like rah rahā (m) hai ‘he is living (m)’, Megha signals that Sulekha
is estranged from her.
3.3 The use of masculine self-reference

Even more interesting than the above uses of the third-person masculine is the
hijras’ use of the first-person masculine. While this rarely surfaces in conversa-
tion, there are at least three situations in which the Banaras hijras do use it. The
first of these is perhaps the most predictable: The hijras will regularly use the
first-person masculine when referring to themselves as boys or telling of their
childhoods. This linguistic shift follows from the fact that many hijras have
what might be called a discontinuous gender identity which gradually changes
from male to non-male after arrival in the hijra community. It is perhaps for
this reason that the hijras sometimes refer to fellow hijras as masculine when
referring to them in a pre-hijra state, such as when they tell of each other’s
childhoods. Here masculine marking will often perform as a tense marker,
suggesting a time period prior to the hijra’s entry into the community. Nanda
(1990: xviii) alludes to similar linguistic shifts in the preface to her ethnography
of the hijras when she explains her translation techniques, pointing out that she
translates pronouns which refer to the hijras as feminine, unless “referring to
the hijra in the past, when he considered himself a male”. A hijra’s use of the
masculine in such instances seems to reflect her own distancing from a previous
self, a self that continuously provides an unpleasant reminder that her feminini-
ty is appropriated instead of genuine.
A second kind of masculine self-reference occurs when hijras wish to add
emphasis to a particular conversational statement. A beautiful example of this
is reproduced in excerpt (14), when Sulekha wants to express her disagreement
Hindi 155
with Megha’s portrayal of the hijra community. As I mentioned earlier, Megha

is very protective of her community’s reputation, and as a result, she painted for
us a rather conservative portrait of the hijra lifestyle. Many hijras supplement
their income through sex-work, particularly now that their traditional role as
the blessers of newborns has become suspect in some segments of the general
population. Many educated middle-class Indians, for instance, no longer believe
in the hijras’ power over procreation, and refuse to pay them money for their
song and dance performances at birth celebrations. But Megha nevertheless
denied any community involvement with prostitution, and insisted throughout
her conversations with us that hijras were nothing but respected religious
ascetics. In the excerpt below, Sulekha contradicts Megha’s claims, punctuating
her opposition with the first-person masculine singular:
(14) “No, I don’t tell lies!” Sulekha 1993
Vinita: kaise kah rahı̄ thı̄ V: Then why was [Megha] saying,
“ham log ko dukh hotā hai, “We have a lot of sadness.
ham log kā parivār nahÖı rahtā, We no longer have a family.
ham log kā sambandh nahÖı rahtā, We no longer have relationships.
ham log bhı̄ sote u»thte bait» hte”. All we do is sleep, get up, sit around”?
Sulekha: nahÖı . ye galat bāt hai. – S: No, that’s wrong.
galat bāt hai. That’s wrong.
maı̃ isko nahÖı māntām. – I don’t believem that.
galat bāt hai. That’s wrong.
admı̄ ke sāth kartā hai sab. – They all have relationships with men.
jaise aurat *mard sambandh hota hai, – *Hijras have relationships with men
» mard ke sāth sambandh
usı̄ tarah *hijre just like women have relationships
hotā hai. with *men.
kitne *hijre-» So many *hijras-
kitne hijre» rakh lete haı̃ ādmı̄ ko, – So many hijras keep men.
kitnā peśāvar hotā hai, (1.0) So many are professionals.
peśā kartı̄, Those who do it as a profession
tab (1.0) *sau, pacās, do sau, cār sau, charge *100, 50, 200, 400,
*sabkā peśā kartı̄ haı̃. – *anything they can get.
maı̃ jhū»th kahtām hu?Ö Do I tellm lies?
nahÖÖı kahtām hu.
Ö No, I don’t tellm lies.
Here, Sulekha also repeatedly employs the infamous hijra flat-palmed clap,
indicated in my transcription system by an asterisk. Clapping five times in this
short passage (and the hijras’ clap, incidentally, is extremely sharp and loud),
Sulekha denies everything Megha has told us: “Yes, we have relationships with
men; yes, we work as prostitutes; yes, we charge anything we can get.” Use of
the first-person masculine helps to bring her point home, and as this is highly
156 Kira Hall
uncharacteristic for Sulekha, she commands our attention: “maı̃ jhū»th kahtā
(m) hu?Ö nahÖÖı kahtā (m) hu!”
Ö ‘Do I tell (m) lies? No, I don’t tell (m) lies!’
A final example of the hijras’ use of the first-person masculine comes from
a hijra community on the outskirts of the city. The four hijras who make up
this comparatively isolated community, all born into Hindu families who
ostracized them, have now adopted the religious practices of the Muslim
families they live beside – families who in many ways suffer a similar marginal-
ization as residents of a city that is thought of throughout northern India as the
“holy Hindu city”. The 80 year old Shashi is the guru of the group, and after 69
years of speaking like a woman (she became a hijra when she was eleven years
old), we rarely heard her use masculine self-reference. But during one visit we
accidentally instigated a family dispute that disrupted this small community.
We often gave gifts to the hijras the first few times we met with them, so as to
express our gratitude for the time they spent with us and show respect for their
profession. On this day, however, we decided to give a gift of 101 rupees to
Shashi’s chela Mohan, since we had given Shashi a colorful saree on our
previous visit. But this is not how the hijra hierarchy works: Only the guru is
allowed to accept gifts, and the guru then distributes the group’s earnings
among her chelas. By giving Mohan a gift, we inadvertently (and regrettably)
upset the community balance. After we left that day, Mohan apparently
refused to hand over the rupees to her guru, and when confronted with
Shashi’s rage, she fled back to her home village outside of Delhi.
When we returned a week later, we immediately sensed that something was
wrong. The hijras’ house was deserted except for Shashi, who we found sitting
on a small cot, slumped against a wall. She was devastated by Mohan’s depar-
ture, and was grieving for the loss of her favorite chela. No apology we gave
could remedy the situation or sufficiently express our regret, so we sat with her
nervously and listened to what she had to say. Wailing merā be»tā, merā be»tā ‘my
son, my son’ and clapping in anger, Shashi screamed about what had happened,
venting her anger entirely through use of the masculine first- and third- person.
It would seem that for Shashi, anger is an emotion best expressed in the
masculine. Perhaps rage is a gut-level reaction that recalls the masculine forms
she produced prior to her entry into the community, or perhaps rage requires
the kind of emphatic masculinity used by Sulekha in the previous passage.
Whatever the case, Shashi’s use of masculine self-reference became a dramatic
and forceful tool for venting this rage, and it remains with me as a painful
reminder of the inherent imbalance between a researcher and her subject.
Hindi 157
4. Conclusions
But what has this discussion of what we might call “unnatural” gender by
Hindi-speaking hijras told us about gender in non-hijra Hindi-speaking
communities? Although the morphological shifting exhibited here is perhaps
unique to the hijra community, I suggest that the conversations of women and
men are also subject to comparable kinds of “gendered negotiations”. As
language and gender theorists have begun to demonstrate (see the articles in
Hall & Bucholtz 1995, for instance), men and women in a variety of communi-
ties exploit cultural expectations of femininity and masculinity in order to
establish positions of power and solidarity. The hijras have an added resource
for accomplishing this, as their between male- and female-position allows them
to access and claim both sides of the grammatical gender divide. But non-hijras
“do gender” in conversational interaction too, working with and against
ideologies of feminine and masculine speech that are themselves rooted in
cultural expectations of gender-appropriate behavior. It is here that Judith
Butler’s (1990, 1993) Derridean reworking of J. L. Austin’s concept of perform-
ativity becomes useful. Her argument that gender works as a performative,
constituting the very act that it performs, leads us away from sociolinguistic
approaches to identity that view the way we talk as directly indexing a predis-
cursive self. For Butler, there is no prediscursive self, as even our understanding
of “biological sex” is discursively produced. This theoretical perspective throws
a decisive wrench in the distinction between “natural gender” and “grammatical
gender”, since there is no “natural” in Butlerian theory. Without the concept of
natural gender to fall back on, sociolinguists can no longer make the circular
claim that speaker X speaks like X because he is male or that speaker Y speaks
like Y because she is female. Rather, we must turn our focus to the speech event
itself, uncovering how speakers manipulate ideologies of femininity and
masculinity in the ongoing production of gender.
In the case of the hijras, we located some of these ideologies in the associa-
tion of mardānā bolı̄ with directness, hierarchy, and anger, and zanānā bolı̄ with
indirectness, solidarity, and intimacy. The Banaras hijras at times challenge such
associations in their creative employments of masculine and feminine self-
reference, but their use of grammatical gender is nevertheless constrained by a
rather traditional and dichotomous understanding of gender. While the hijras
tend to use the masculine when speaking to a superior or inferior, emphasizing
particular points, recalling their past selves as non-hijras, or expressing intense
anger, they are more likely to employ the feminine when expressing solidarity
158 Kira Hall
and intimacy with fellow community members. Because they occupy an

ambiguous gender position in a city that continues to marginalize them,
Banaras hijras are perhaps more attentive to the role speech plays in the
performance of gender. With a heightened awareness of how language can
index gender identity, they enact and contest ideologies of gendered speech in
their daily interactions. The research is not yet here to tell us whether “mono-
sexed” Hindi-speaking communities in India share these same expectations of
male and female verbal behavior; nor do we yet know how such expectations
might influence the actual language practices of men and women in specific
communities of practice. I offer this analysis of a rather unusual community to
excite more research on gender in Hindi sociolinguistics. The language practices
of these Banaras hijras, extraordinary as they are, are not created in a cultural
vacuum. Their discursive choices are influenced by dominant ideologies of
gender in northern India – ideologies that no doubt affect linguistic behavior in
a variety of Hindi-speaking communities.
Appendix
The transcription conventions used for the Hindi passages in this article include the
following; I have not used all of these conventions in the English translations since extra-
linguistic features like intonation and emphasis are not parallel.
xf superscripted f indicates feminine morpho- [] brackets enclose words added to clarify

logical marking in the Hindi the meaning of the text
ym superscripted m indicates masculine mor- what bold print indicates syllabic stress
phological marking in the Hindi : a colon indicates a lengthening of a
(0.5)indicates length of pause within and between sound (the more colons, the longer the
utterances, timed in tenths of a second sound)
a – a a dash with spaces before and after indicates a . a period indicates falling intonation
short pause, less than 0.5 seconds , a comma indicates continuing intona-
but- a hyphen immediately following a letter tion
indicates an abrupt cutoff in speaking ? a question mark indicates rising intona-
(( )) double parentheses enclose nonverbal move- tion at the end of a syllable or word
ments and extralinguistic commentary … deletion of some portion of the original
text
“a” quotation marks enclose quoted or
reported speech
* flat-palmed clap characteristic of the
hijras
< > overlapping talk
<DEST "hal-n*">
Hindi 159
Notes
* I began this field research in a joint project with Veronica O’Donovan after concluding an
advanced language program during the 1992–1993 academic year in Banaras, India. I am
grateful to the American Institute of Indian Studies for sponsoring my participation in the
program. Some of the data discussed in this essay also appears in Hall & O’Donovan (1996).
My thanks are extended to the many people who helped me with this project in both India
and America, to my friend and Hindi teacher Ved Prakash Vatuk, and to the Banaras hijras
who participated in these discussions.
1. This percentage is taken from the 1991 Census of India, in which 4022 of every 10,000
people listed Hindi as their primary language and another 518 listed Urdu. These and other
language statistics are available at the Census of India website: www.censusindia.net.
2. See Kachru (1980:3–11) for an interesting discussion of the grammatical tradition of Hindi.
3. There are, however, a number of insightful studies on women’s uses of various South
Asian oral traditions, such as Raheja & Gold’s (1994) ethnography on the poetics of women’s
resistance in the songs, stories, and personal narratives of women in northern India.
4. For a more extensive discussion of Valentine’s work, see Hall (forthcoming).
5. See Romaine (1999: 63–90) for a recent discussion of the “leakage” between grammatical
gender and natural gender.
» bolı̄ area is generally said to be located in western Uttar
6. The standard Hindi kharı̄
Pradesh between the Ganges and Jumna rivers, stretching from Dehradun in the north to
Bulandshahar in the south.
7. In Delhi the hijras have named their language Farsi. While their “hijralect” has very little,
if anything, to do with what is generally known as Farsi, the term is fitting given that the
hijras see themselves as descended from the eunuchs of the medieval Moghul courts, where
Farsi was the dominant language.
»
8. I am grateful to Ved Vatuk for bringing to my attention the parallel between sadhukkarı̄
» bolı̄.
and hijrā
9. I use the term subject in the sense of syntactic subject, not morphological subject. The
syntactic subject may be in an oblique case form and does not necessarily determine
agreement. The term is usually used for the NP which corresponds to the subject of the
English translation.
10. In recent years, however, parents in some Hindi-speaking areas have begun to use the
feminine term be»tı̄ in direct address to a son, also for reasons of endearment (Ved Vatuk,
personal communication).
11. The term bhāı̄, like the masculine term yār ‘friend’, is used in a variety of contexts to
index informality. Wives, for instance, will sometimes refer to their husbands as bhāı̄.
12. Ved Vatuk (personal communication).
13. Except, of course, in theatrical situations when men play women’s roles, as in folk-
dramas known as nautankı̄ or sāng.
160 Kira Hall
14. A similar reaction would most likely occur towards women using masculine self-
reference, except for the fact that there has recently been a number of public female figures
who have challenged the assumption of “natural” linguistic gender by assuming a male
speaking voice. Most notable in this regard is one of the lead characters in the sitcom Ham
Panch, whose use of the masculine first-person is intended to reflect her tomboyish
personality. Moreover, the popular singer Milan Singh, who is known for singing both male
and female parts of traditional Bollywood film songs, regularly speaks in the first person
plural when interviewed so as to keep her gender ambiguous before the public.
15. I have chosen pseudonyms for all of the hijras appearing in this article and have avoided
giving the names of the four hijra communities mentioned to protect their anonymity. I have
also chosen to use her and she to refer to the hijras since they prefer to be referred to and
addressed in the feminine.
16. It could also be quite possible that the group of people Sulekha is referring to here are
actually jankhās (as called in Banaras) or kotı̄s (as called in Delhi), i.e. men who dress and
dance as women but do not publicly identify as hijras or officially belong to the hijra
community. As these “non-castrated” men continue to remain in their extended family
structures, undergoing arranged marriages and bearing children, they sometimes receive a
certain amount of animosity from the hijras. See Hall (forthcoming) for a more lengthy
discussion of these different groups.
References
Apte, Mahadev L. 1974. “Pidginization of a Lingua Franca: A linguistic analysis of Hindi-

Urdu spoken in Bombay.” Contact and Convergence in South Asian Languages. Interna-
tional Journal of Dravidian Linguistics 3: 21–41.
Beames, John. 1872–79. A comparative grammar of the modern Aryan languages of India. 3
vols. London.
Bhatia, Tej K. 1992. “Variation in Hindi: Parameters for a pan-dialectal grammar.” In
Dimensions of sociolinguistics in South Asia: Papers in memory of Gerald B. Kelley, eds.
Edward C. Dimock & Braj B. Kachru & Bhadriraju Krishnamurti. New Delhi: Oxford
& IBH Press, 163–197.
Bloch, Jules. 1910. “Castes et dialectes en Tamoul.” Mémoires de la Société de Linguistique
de Paris 16: 1–30.
Bodine, Ann. 1975. “Androcentrism in prescriptive grammar: Singular ‘they’, sex-indefinite
‘he’, and ‘he or she’.” Language in Society 4: 129–146.
Börner-Westphal, Barbara. 1989. “Sexus-Genus-Beziehungen im Hindi.” Zeitschrift für
Phonetik, Sprachwissenschaft und Kommunikationsforschung 42: 762–770.
Butler, Judith. 1990. Gender trouble: Feminism and the subversion of identity. New York:
Routledge.
Butler, Judith. 1993. Bodies that matter: On the discursive limits of “sex”. New York: Routledge.
Chatterji, Suniti Kumarr. 1931. “Calcutta Hindustani: A study of a jargon dialect.” Indian
Linguistics 1: 2–4. Reprinted 1972 in Select papers. New Delhi: People’s Publishing
House, 204–256.
Hindi 161
Chernyshev, V. A. 1971. “Nekotorye cherty Bombeiskogo Govora Khindustani, (na materiale

sovremennoi prozy khindi)” [Some features of Bombay Hindustani speech (in modern
Hindi prose material)]. In Indiiskaia i Iranskaia Filologiia (Voprosy Dialektologii), ed.
Nikolai Dvoriankov. Moscow: Izdatel’stvo “Nauka”, 121–141.
Gosvāmı̄, Pūranagira. 1979. Hindı̄ mem» linga-bheda kā adhyayana [A study of gender
distinctions in Hindi]. Varanasi: Sañjaya Prakāśana.
Grierson, George Abraham. 1903–1928. Linguistic survey of India (1903–28). 11 vols. Reprint
1967. Delhi: Motilal Banarsidass.
Gumperz, John J. 1958. “Dialect differences and social stratification in a North Indian
village.” American Anthropologist 60: 668–682.
Guru, Kamta Prasad. 1920. “Hindı̄ Vyākaran» ” [Hindi grammar]. Varanasi: Kashi Nagri
Pracharini Sabha.
Hall, Kira. 1997. “‘Go suck your husband’s sugarcane!’ Hijras and the use of sexual insult.”
In Queerly phrased: Language, gender, and sexuality, eds. Anna Livia & Kira Hall. New
York: Oxford University Press, 430–460.
Hall, Kira (forthcoming). From tooth to tusk: Language and gender in an Indian hijra
community. Unpublished manuscript.
Hall, Kira & Mary Bucholtz, eds. 1995. Gender articulated: Language and the socially con-
structed self. New York: Routledge.
Hall, Kira & Veronica O’Donovan. 1996. “Shifting gender positions among Hindi-speaking
Hijras.” In Rethinking language and gender research: Theory and practice, eds. Victoria L.
Bergvall & Janet M. Bing & Alice F. Freed. London: Longman, 228–266.
Jain, Dhanesh Kumar. 1969. “Verbalization of respect in Hindi.” Anthropological Linguistics
11: 79–97.
Jain, Dhanesh Kumar. 1973. Pronominal usage in Hindi: A sociolinguistic study. Ph.D.
dissertation, University of Pennsylvania.
Kachru, Yamuna. 1980. Aspects of Hindi grammar. New Delhi: Manohar Publications.
Kellog, Samuel H. 1875. A grammar of the Hindi language. London: Routledge.
Khubchandani, Lachman M. 1978. “Towards a selection grammar: Fluidity in modes of
address and reference in Hindi-Urdu.” Indian Linguistics 39: 1–4.
Khubchandani, Lachman M. 1991. “India as a sociolinguistic area.” Language Sciences 13:
265–288.
King, Christopher R. 1999. One language two scripts: The Hindi movement in nineteenth
century northern India. New York: Oxford University Press.
Lakoff, Robin. 1973. “Language and woman’s place.” Language in Society 2: 45–80.
Martyna, Wendy. 1980a. “Beyond the he/man approach: The case for nonsexist language.”
Signs: Journal of Women in Culture and Society 5: 482–493.
Martyna, Wendy. 1980b. “The psychology of the generic masculine.” In Women and
language in literature and society, eds. Sally McConnell-Ginet & Ruth Borker & Nancy
Furman. New York: Praeger, 69–78.
McGregor, Ronald Stuart. 1972. Outline of Hindi grammar. Delhi: Oxford University Press.
Mehrotra, Raja Ram. 1977. “Fluidity in kinship terms of address.” Anthropological Linguistics
19: 123–132.
Mehrotra, Raja Ram. 1985a. Sociolinguistics in Hindi context. New Delhi: Mohan Primlani.
(“Hindi address forms”, 39–79)
</TARGET "hal">
162 Kira Hall
Mehrotra, Raja Ram. 1985b. Sociolinguistics in Hindi context. New Delhi: Mohan Primlani.
(“Modes of greeting in Hindi: A sociolinguistic statement”, 80–122)
Mehrotra, Raja Ram. 1985c. “Sociolinguistic surveys in South Asia: An overview.” Interna-
tional Journal of the Sociology of Language, 55: 115–124.
Misra, K. S. 1977. Terms of address and second person pronominal usage in Hindi: A socio-
linguistic study. New Delhi: Bahri Publications.
Mohanan, Tara. 1994. Argument structure in Hindi. Stanford: CSLI Publications.
Nanda, Serena. 1990. Neither man nor woman: The hijras of India. Belmont, CA: Wadsworth.
Nespital, Helmut. 1990. “On the relation of Hindi to its regional dialects.” In Language versus
dialect: Linguistic and literary essays on Hindi, Tamil, and Sarnami, ed. Mariola Offredi.
New Delhi: Manohar Publications, 3–23.
Pathak, Anand Swarup. 1976. Hindı̄ kā linga vidhāna [On gender in Hindi: A study].
Mathura: Pramod Prakāśan.
Raheja, Gloria Goodwin & Ann Grodzins Gold. 1994. Listen to the Heron’s words: Reimagining
gender and kinship in North India. Berkeley, CA: University of California Press.
Romaine, Suzanne. 1999. Communicating gender. Mahwah, NJ: Erlbaum.
Sachdeva, Rajesh. 1982. Sociolinguistic profile of Milijhuli around Delhi. Ph.D. Dissertation,
Poona University, India.
Shapiro, Michael C. 1989. A primer of modern standard Hindi. Varanasi: Motilal Banarsidass.
Shapiro, Michael C. & Harold F. Schiffman. 1983. Language and society in South Asia.
Dordrecht: Foris.
Sharma, Aryendra. 1958. A basic grammar of Hindi. Delhi: Government of India.
Simon, Beth Lee. 1993. “Language choice, religion, and identity in the Banarsi community.”
In Living Banaras: Hindu religion in cultural context, eds. Bradley R. Hertel & Cynthia
Ann Humes. Albany: State University of New York Press, 245–268.
Simon, Beth Lee. 1996. “Gender, language, and literacy in Banaras, India.” In Gender and
belief systems: Proceedings of the Fourth Berkeley Women and Language Conference, eds.
Natasha Warner & Jocelyn Ahlers & Leela Bilmes & Monica Oliver & Suzanne Wert-
heim & Mel Chen. Berkeley: Women and Language Group, 679–687.
Simon, Beth Lee. Glass River. Unpublished manuscript.
Singh, Rajendra & Jayant K. Lele. 1990. “Language, power, and cross-sex communication
strategies in Hindi and Indian English revisited.” Language in Society 19: 541–546.
Srivastava, Murlidhar. 1969. The elements of Hindi grammar. Varanasi: Motilal Banarsidass.
Vajpeyi, Kishoridass. 1967. Hindii shabdaanushaasan [Hindi grammar]. Varanasi: Kashi
Nagri Pracharini Sabha.
Vajpeyi, Kishoridass. 1968. Acchii Hindi [Good Hindi]. Meerut: Minakshi Prakashan.
Valentine, Tamara M. 1985. “Sex, power, and linguistic strategies in the Hindi language.”
Studies in the Linguistic Sciences 15: 195–211.
Valentine, Tamara M. 1986. “Language and power: Cross-sex communicative strategies in
Hindi and Indian English.” Economic Political Weekly (Special Issue: Review of Women
Studies), 75–87.
Valentine, Tamara M. 1987. “Sexist practices in the Hindi language.” Indian Journal of
Linguistics 14: 25–55.
Vatuk, Sylvia. 1969a. “Reference, address, and fictive kinship in urban North India.”
Ethnology 8: 255–272.
Vatuk, Sylvia. 1969b. “A structural analysis of the Hindi kinship terminology.” Contributions
to Indian Sociology, New Series 3: 94–115.
<LINK "gun-n*">
<TARGET "gun" DOCINFO
AUTHOR "Anna Gunnarsdotter Grönberg"
TITLE "Masculine generics in current Icelandic"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
icelandic
Masculine generics in current Icelandic*
Anna Gunnarsdotter Grönberg

University of Göteborg, Sweden
1. Introduction
2. Gender in the linguistic structure of Icelandic
2.2 Grammatical gender and personal reference
2.3 Masculine generics
2.4 Occupational titles for women
3. Language and gender in Iceland
4. The use of Icelandic occupational titles: A pilot study
5. Conclusion
Notes
References
1. Introduction
Icelandic (Íslenska) is a North Germanic language closely related to Faroese,

Norwegian, Swedish, and Danish. Iceland was settled by Norwegians in the 9th
century. The country was self-governed during the first centuries, but from the
middle of the 13th century it was ruled by the Norwegian king, and later by the
Danish king, when Norway became part of Denmark in the 15th century. In
1944 Iceland became an independent republic. During the time of Danish rule,
Danish was the language used by authorities (Haugen 1976: 351), but Icelandic
appears to have been used in all other domains. During the centuries of self-
government and Norwegian rule, a literary tradition had flourished which
produced texts such as the Family Sagas, and the language had developed from
Old Norse (Norwegian) into what is now recognized as Icelandic. Icelandic was
used for writing during the Danish rule (Vikør 1995: 57), and the Icelandic
164 Anna Gunnarsdotter Grönberg
translation of the Bible, published after the Lutheran Reformation in the 16th
century, was of great importance for the status of Icelandic. Danicisms and
semi-Danish orthography were combated successfully during the 18th and 19th
centuries (Haugen 1976: 398).
Icelandic underwent extensive phonetic changes during the Middle Ages,
and during the 20th century the vocabulary has expanded dramatically. What
remains of the earlier stages of Icelandic is mainly the orthography, the mor-
phological and syntactic structures, and the basic vocabulary (Iceland 1986:59f).
Icelanders have cultivated a tradition of puristic language planning, and when
modern times require new words, e.g. technical terminology, loanwords are
mostly rejected and new words coined from existing indigenous elements, or
old words are given a new meaning. Since Iceland is a small nation with
approximately 280,000 inhabitants (cf. Year Book 2000: 1768), and since it was
under foreign rule for hundreds of years, the arguments for linguistic purism
are in many ways nationalistic: “[…] drastic changes in the structure and/or the
vocabulary of the language would estrange Icelanders from their cultural
heritage and thus fatally weaken their sense of being a separate people” (Vikør
1995: 210–211).
2. Gender in the linguistic structure of Icelandic

Icelandic is a mainly synthetic language where inflectional suffixes are attached
to stems, marking linguistic categories such as tense, mood, person, number,
gender, or case (cf. Jónsson 1927, Einarsson 1945, Kress 1982, and Svavars-
dóttir & Jónsdóttir 1998). In addition to suffixes, grammatical categories can
also be realized by zero morphemes or vowel changes in the root, i.e. by
umlaut and ablaut.
Nouns in Icelandic have grammatical gender: masculine, feminine or
neuter. Nouns are inflected for number (singular and plural) and case (nomina-
tive, genitive, dative and accusative). Depending on gender and case endings,
nouns can be divided into two main inflectional classes, the so-called strong and
weak declensions. Since Icelandic is a strongly inflectional language, there is a
vast number of word forms. However, besides some irregular forms, most of the
inflectional paradigms share a large amount of features. Examples of common
morphemes are -ur, -inn (strong) or -i (weak) for masculine singular nomina-
Icelandic 165
tive, -Ø, -in (strong) or -a (weak) for feminine singular nominative, and -Ø, -ið,
-t (strong) or -a (weak) for neuter singular nominative. Apart from inflectional
endings, there is an umlaut (a>e, a>ö) in some cases in which the ending is zero
or contains /i/, /u/.
Pronouns (except relative pronouns),1 free definite articles, adjectives, past
participles,2 and certain numerals (ordinals and the cardinals from one to
four) must agree in gender, case and number with their nouns (or referents, if
there is no noun present in the co-text). This is the case when they are in a
construction with a noun within a noun phrase, if they are used anaphorically,
if they are predicative complements, and if they act independently as subjects.
Examples (1a–f) show gender and number agreement of nouns, pronouns,
adjectives, and numerals.
(1) a. Þessi strák-ur er
this.masc.sg.nom boy-masc.sg.nom is
svang-ur.
hungry-masc.sg.nom
‘This boy is hungry.’
b. Þessi stelp-a er svöng.
this.fem.sg.nom girl-fem.sg.nom is hungry.fem.sg.nom
‘This girl is hungry.’
c. Þetta barn-Ø er svang-t.
this.neut.sg.nom child-neut.sg.nom is hungry-neut.sg.nom
‘This child is hungry.’
d. Tveir strák-ar eru svang-ir.
two.masc.nom boy-masc.pl.nom are hungry-masc.pl.nom
‘Two boys are hungry.’
e. Tvær stelp-ur eru svang-ar.
two.fem.nom girl-fem.pl.nom are hungry-fem.pl.nom
‘Two girls are hungry.’
f. Tvö börn eru svöng.
two.neut.nom child.neut.pl.nom are hungry.neut.pl.nom
‘Two children are hungry.’
2.2 Grammatical gender and personal reference

Gender assignment in Icelandic is considered to be arbitrary. Even animate
nouns denoting living creatures or human beings do not necessarily show a
correspondence between grammatical gender and referential gender.
In two standard grammars of Icelandic, Guðmundsson (1922) and Kress

(1982), the three-gender system is taken for granted and no comments are
made upon how or why nouns are divided into three gender classes. In two
major university textbooks the gender system is presented as “mostly grammati-
cal” (Einarsson 1945: 32) and “principally grammatical” (Friðjónsson 1978: 16),
as opposed to what they call the natural gender of English, but no further
comments are made upon gender assignment. Jónsson (1927:14) writes that “in
the case of names or appellations of persons and animals the gender is generally
determined by the natural sex. Otherwise it mainly depends upon the termina-
tion of the noun, but universal rules for all cases cannot be given.” In a more
recent textbook, Svavarsdóttir & Jónsdóttir (1998) discuss the relationship
between grammatical gender and referential gender:
The gender of words is grammatical, it depends on the form of a word rather
than its meaning. The gender of a word and sex of its referent need not
necessarily correspond. It is possible to use a masculine word when referring to
a woman, a feminine word when referring to a man, and words used for living
creatures – either people or animals – could be neuter. However, there is
mostly a correspondence between sex and gender when a word refers to living
individuals. Words that refer to inanimate objects, thoughts or other things
that are sexless, can be masculine or feminine, just as well as neuter.
(Svavarsdóttir & Jónsdóttir 1998: 3, my translation)
Table 1 illustrates the partial correspondence between grammatical gender and

referential gender in personal nouns (cf. Svavarsdóttir & Jónsdóttir 1998: 3). As
can be seen here, the most frequent nominative endings of personal nouns are
-ur and -i for masculine, -a for feminine, and -Ø and -i for neuter.
Most kinship terms are inherently gender-specific with agreement between
grammatical and lexical gender. Examples of feminine/females: móðir ‘mother’,
systir ‘sister’, frænka ‘female cousin or relative; aunt’; masculines/males: faðir
‘father’, bróðir ‘brother’, frændi ‘male cousin or relative; uncle’. One exception
is ‘parent/s’, which is neuter in the singular (foreldri) and masculine in the
plural (foreldrar).
In the area of other common nouns there are a number of examples which
contradict a straightforward relationship between masculine/male and femi-
nine/female. Examples are: kvenmaður (m) ‘woman’, kvenkostur (m) ‘nice
woman, good match (for a man)’, skáld (n) ‘poet’, morðkvendi (n) ‘murderess’
mannskepna (f) ‘evil man, mankind’, hetja (f) ‘hero’ and lögga (f) ‘cop’ (there
are still very few female police officers in Iceland). Kvenmaður and kvenkostur
are compounds with kven ‘female’, which comes from the same root as kona
Icelandic 167
Table 1.The gender of personal nouns

Masculine Feminine Neuter
maður ‘man, human’ manneskja ‘human’ góðmenni ‘altruist’

einstaklingur ‘individual’ persóna ‘person’ illmenni ‘villain’
karl-maður ‘man’ kona ‘woman’ ungmenni ‘youth’
karl ‘man, old man’ kerling ‘old woman’ barn ‘child’
kven-maður ‘woman’ mæðg-ur (pl) ‘mother and mæðg-in (pl) ‘mother
daughter/s’ and son/s’
feðg-ar (pl) ‘father and son/s’ stelpa ‘girl’ feðg-in (pl) ‘father and
daughter/s’
foreldr-ar (pl) ‘parents’ stúlka ‘girl’ foreldri ‘parent’
strákur ‘boy’ mey ‘maid’ hjón ‘couple’
drengur ‘boy’ mær ‘maid’
krakki ‘kid’ kennslu-kona ‘woman teacher’
táningur ‘teenager’ hjúkrunar-kona ‘(woman) nurse’
kennari ‘teacher’ skáld-kona ‘poetess’ skáld ‘poet’
hjúkrunar-fræ- ‘nurse’ hetja ‘hero’
ðingur
læknir ‘physician’
höfundur ‘author’
sjúklingur ‘patient’
landi ‘compatriot’
(ríkis)borgari ‘citizen’
bílstjóri ‘driver’
viðskiptavinur ‘customer’
nemandi ‘student’
þáttakandi ‘participant’
‘woman’, but the female denotation of the compounds does not change their
gender assignment, since their base lexemes maður ‘man, human’ and kostur
‘choice’ remain masculine. The use of feminine nouns with male reference is
rather uncommon, though (examples are hetja and lögga, feminine epicenes,
which are most likely to have male referents); there are far more masculine
nouns referring to women than vice-versa.
In cases where grammatical and referential gender do not correspond,
grammatical gender determines the choice of the personal pronoun. When
talking about a poet, for instance, as skáld (n), without mentioning any particu-
lar female or male poet, it is correct to refer to her/him with það ‘it’. Such is also
the case with the slang word lögga (f) ‘cop’, which requires feminine anaphoric
pronouns such as hún ‘she’. However, in contexts that refer to a specific male
individual, a masculine pronoun may be used. Semantics then overrides
grammatical agreement. Similarly, when Vigdís Finnbogadóttir was president
of Iceland (1980–1996), she received the masculine title forseti (m) ‘president’,
which requires the male personal pronoun hann ‘he’. However, the female
pronoun hún ‘she’ could also be used, resulting in a choice between feminine
agreement (referring to Vigdís Finnbogadóttir) and masculine agreement
(referring to forseti).3 In written language, examples of this kind are not very
frequent, since language users seem to avoid constructions where they should
refer to a person with a pronoun that contradicts referential gender. In a
questionnaire, Theódórsdóttir (1985) observed that informants disfavored
constructions of this kind:
[…] there is a strong tendency that sex determines the gender assignment of a
personal pronoun or the form of a verb [i.e. participles], which refers to the
subject, if there is a disagreement between the gender of the subject and the sex
of the referent. Especially if something within the sentence reveals the sex of
the person to whom the subject is referring.
(Theódórsdóttir 1985: 39, my translation)
There were no clear examples of disagreement between gender of a pronoun

and referential gender in the analyzed issue of the newspaper Morgunblaðið (cf.
Section 4), but such examples can be found in other sources. In example (2),
taken from a novel, a neuter pronoun refers to a prince, when he is referred to
by a neuter noun, keisaraefni ‘emperor-to-be’:
(2) þau (n.pl) höfðu hist í skóginum þegar kerlingin (f.sg) var ung stúlka og
keisarinn (m.sg) óbreyttur prins. Keisaraefnið (n.sg) hafði verið á þrasta-
veiðum með mönnum sínum þegar það (n.sg) kom auga á ref sem […]
‘They (n.pl) had met in the forest when the old woman (f.sg) was a
young girl and the emperor (m.sg) was only a prince. The future emper-
or (n.sg) had been hunting thrushes with his (‘its’) men when he (‘it’
n.sg) saw a fox that […]’
(Sjón 1994: 75, my translation)
In example (3), the writer makes different choices of gender agreement when
referring to a group of interviewed school nurses. They are never mentioned by
their names, only referred to by the masculine noun (skóla-)hjúkrunarfræðingur
‘(school) nurse’, but it is also stated that they are all women. The first sentence
of example (3) is comparable with the second sentence of example (2). In both
cases a noun (keisaraefni/hjúkrunarfræðingur) of a gender different from the
gender of its referent is introduced in the first clause. In example (2), the
anaphoric pronoun (það) agrees grammatically with the neuter noun. In
example (3), however, the anaphoric pronoun (þær) agrees semantically with
the gender of its female referents:
Icelandic 169
(3) Ályktanir mínar af þessum viðtölum við hjúkrunarfræðinga (m.pl) er að

þær (f.pl) eru mjög áhugasamar (f.pl) um störf sín en kvarta margar (f.pl)
yfir […] Flestir (m.pl) hjúkrunarfræðingarnir (m.pl) tala um tímaskort og
sumir (m.pl) segja […]
Flestar (f.pl) bentu þá á að sú samræming sem verið er að gera um störf
skólahjúkrunarfræðinga (m.pl) verði mun til bóta. það er mitt mat að
hjúkrunarfræðingar (m.pl) séu mjög til þess fallnir (m.pl) að […].
‘My conclusions from these interviews with nurses (m.pl) is that they
(f.pl) are very interested (f.pl) in their work, but many (f.pl) of them
complain about […] Most (m.pl) of the nurses (m.pl) talk about lack of
time, and some (m.pl) of them say […]
Most (f.pl) of them emphasized that the new coordination of the work of
school nurses (m.pl) will improve the situation. It is my opinion that
nurses (m.pl) are well suited (m.pl) for […]’
(Bjarnadóttir 1996: 19, my translation)
When comparing the different sentences of example (3), gender agreement

appears to depend on the position of the pronoun or adjective within the clause.
Predicative (áhugasamar, f; fallnir, m) and attributive (flestir, m) adjectives
agree grammatically with their headwords. The adjectives margar (f) and flestar
(f), which are in independent subject positions and not in the same clause as the
masculine noun, show semantic referential agreement with the text-external
individuals. Accordingly, semantic agreement is possible when an adjective
holds an independent position, but not when it is predicative or attributive.
When a pronoun has an independent position, gender agreement seems to
depend on the sort of pronoun. The personal pronoun þær (f) ‘they’ agrees
semantically with the referents’ gender instead of with the preceding noun, but
the indefinite pronoun sumir (m) ‘some’ agrees with the grammatical gender of
hjúkrunarfræðingur (m) ‘nurse’. This can be compared to example (5) in
Section 2.3. Indefinite pronouns, which will be further discussed in a later
section, are often generically masculine. Even if sumir holds an independent
position similar to þær and flestar and refers to the group of interviewed women
(which becomes clear from the wider co-text), the generic masculine is chosen.

The fact that grammatical and referential gender do not always correspond in
current language use leads to another aspect of grammatical gender – neutrality,
or the question which gender is unmarked within a given system, i.e., a linguistic
system as a whole or language use in a particular context. This aspect of gender

has, as far as I can tell, hardly been discussed in Iceland until recently. Rögn-
valdsson (1990: 62; first published in 1983) seems to be the first to treat marked-
ness of gender, number and case in an educational book on Icelandic morpho-
logy for university students:
There are three gender classes, masculine, feminine and neuter, and we may
most likely regard neuter as unmarked; it appears in adjectival predicates if the
subject is a sentence (Að María skuli elska Jón er ótrúlegt [n]) ‘That Mary could
love John is incredible’ or in an oblique case (Stelpunni [dative; an impersonal
construction] er kalt (n)) ‘The girl is cold’. Neuter also appears when genera
are mixed; hann + hún ‘he + she’ is neither þeir ‘they’ (m) nor þær ‘they’ (f),
but þey ‘they’ (n).
(Rögnvaldsson 1990: 62, my translation; boldface, except in first line, is mine.)
In an analysis of current Icelandic noun inflection, Svavarsdóttir (1993a; first

published in 1987) comments that:
Greenberg […] [1966: 75] mentions on the one hand examples of references to
a heterogeneous group or gathering, e.g., people of both sexes, and assumes
that the plural of the unmarked category is generally applied in those cases. In
connection with this I could call attention to the neuter plural which is applied
in such circumstances in Icelandic (þau [n.pl] eru farin [n.pl] ‘They are gone’;
Komið þið [pl] sæl [n.pl] ‘How do you do?’, lit. ‘come you blessed’) and, in
fact, Eiríkur Rögnvaldsson (1986: 69–70 [1990: 62–63]) supposes, by similar
arguments, that neuter is the most unmarked gender.
(Svavarsdóttir 1993a: 33, my translation; boldface is mine.)
Rögnvaldsson and Svavarsdóttir come to the conclusion that the neuter is the
most unmarked gender, feminine the most marked and masculine somewhere
in between. This position is controversial, however. There is sufficient evidence
that, on the contrary, masculine is used as the most unmarked gender in most
contexts in modern Iceland, at least as concerns references to human beings. In
fact, Rögnvaldsson also calls attention to the fact that masculine pronouns often
refer to both men and women in generic statements and indefinite pronouns
such as Enginn (m.sg) má yfirgefa húsið ‘Nobody (m.sg) is allowed to leave the
house’ and Allir (m.pl) græða á verðbólgunni ‘Everybody (m.pl) profits from
inflation’ (examples from Rögnvaldsson 1990: 63).
In addition, the masculine is not only unmarked with respect to the
feminine when referring to humans, but also with respect to the neuter. This
conclusion comes from an observation of the widespread use of the masculine
Icelandic 171
in generic statements like those with enginn and allir above. Even though the
neuter could be used when referring to a person of unknown gender or a
heterogeneous group of people, the masculine is normally applied in such cases.
Clear examples are ten advertisements about meetings with Christian congrega-
tions, which appeared in Morgunblaðið (January 29, 1995: B 24). Nine churches
wish ‘everybody welcome’ with a masculine pronoun, Allir (m.pl) velkomnir
(m.pl), and only one uses the neuter form, Öll (n.pl) velkomin (n.pl). There is
a semantic difference between these sentences: Allir velkomnir is generic, while
Öll velkomin has an individualizing function.
There seems to be an unwritten “rule” of language use which is very
productive among speakers of Icelandic: The neuter plural can only be used for
personal and demonstrative pronouns when referring to a number of persons
who are known (and preferably recently mentioned) in the context. When
referring to a group of unspecified individuals or making a generic statement,
the masculine plural is normally used. This is especially important for indefinite
pronouns, where the masculine is to be preferred even when referring to a
group of known individuals (see examples (3) above and (5) below).
Support for the first part of the unwritten “rule” can be found in many
places. In a study on number, case and gender agreement with composite
subjects, Friðjónsson (1990/1991) shows that there is great variation in current
practices of agreement with abstract and uncountable nouns, but when two or
more concrete and countable nouns are composite within a noun phrase, the
gender and number of the predicative complements are neuter plural in most
cases: Konan (f.sg) og barnið (n.sg) eru bæði (n.pl) veik (n.pl). ‘The woman (f.sg)
and the child (n.sg) are both (n.pl) ill (n.pl).’ (Friðjónsson 1990/1991:89).
Svavarsdóttir & Jónsdóttir (1998:11) give a rule for adjectival inflection, when
referring to two or more nouns of different gender: “Note that when the nouns
are not of the same gender, when they are heterogeneous, then the adjective will
mostly be in the neuter plural” (my translation), illustrated by the following
example: Óli (m) og Elsa (f) eru ung (n.pl). ‘Óli (m) and Elsa (f) are young (n.pl).’
The same rule and similar examples are given by Jónsson (1927:14) and Einars-
son (1945:133). Jónsson and Einarsson are also quoted by Corbett (1991:283),
who writes that “neuter plural is used for any mixture of gender”.
The second part of the unwritten rule above, regarding the usage of
masculine plural indefinite pronouns, is not discussed in grammars and
textbooks of Icelandic (cf. Guðmundsson 1922, Jónsson 1927, Einarsson 1945,
Kress 1982, Friðjónsson 1978, Svavarsdóttir & Jónsdóttir 1998), but neverthe-
less grammarians seem to act according to this unwritten rule. Since the issue
is not discussed, the word “rule” may be an overstatement, but the usage of
generic masculine indefinite pronouns is as regular as if there were a rule. When
describing the usage of indefinite pronouns, the grammars mentioned give no
advice for the use of gender when referring to two or more persons of different
or unspecified gender. But the examples given illustrate the observation that
when a noun is present in the sentence, the gender of the pronoun is in agree-
ment with the noun, whereas when pronouns are used generically and have an
independent subject position, the gender chosen is always masculine. Cf. the
following examples:
(4) a. Sumir (m.pl) hafa þann sið að tala við sjálfa (m.pl) sig.
‘Some people have the habit of talking to themselves.’
(Jónsson 1927: 80, my translation)
b. Það eru ýmist allir (m.pl) eða enginn (m.sg) heima.
‘Sometimes everybody is at home, sometimes no one is.’
(Svavarsdóttir & Jónsdóttir 1998: 153, my translation)
Example (5) illustrates how this rule can be realized in a wider context. The
neuter plural personal pronoun þau ‘they’ is only used when referring directly
to the two parents, while the indefinite masculine plural pronoun allir ‘every-
body’ refers to the whole family:
(5) Í dag er laugardagur og þau (n.pl) Páll (m) og Dísa (f) ákveða aðfara í
ferðalag með börnum sínum (n.pl) […] þau (n.pl) aka eftir sléttum mal-
bikuðum vegi og allir (m.pl) eru í góðu skapi. Á leiðinni aka þau (n.pl)
framhjá […]
‘Today it is Saturday and they (n.pl), Páll (m) [father] and Dísa (f)
[mother], decide to set out on a trip with their children (n.pl) [a daugh-
ter and two sons] […] They (n.pl) [Páll and Dísa] drive along a smooth
asphalted road and everybody (m.pl) [the whole family] is in a good
mood. On their way they (n.pl) [Páll and Dísa] drive past […]’
(Sigmundsson 1998: 34, my translation)
Corbett (1991: 218f) does not comment on Icelandic specifically in his discus-
sion of agreement conflicts. But he asserts that a common strategy in Indo-
European languages is to choose masculine agreement when the gender of the
referent is not known (Corbett 1991: 219), which is well in accordance with
Icelandic use of the generic masculine.
Icelandic 173
2.4 Occupational titles for women

An argument for the unmarkedness of the masculine is that there are very few
feminine words that can be used to refer to men and relatively few neuter words
denoting human beings, but a large number of masculine words which can refer
to women. Many of these are occupational titles, like leikari ‘actor’, prestur
‘priest’ and verkamaður ‘worker, workman’. A few female-marked equivalents
with -kona or kven- (leikkona, kvenprestur, verkakona) do exist, but mostly
women have masculine job titles when the original title is masculine.
The most productive patterns for the formation of feminine human nouns
are compounds with -kona and kven-. Kona- is the word for ‘woman’, and kven-
comes from the same root (cf. kven-na, pl. gen. of kona). These two elements
have a similar function and meaning. E.g., words beginning with kven- can also
denote something intended for or pertaining to women, such as kvenfatnaður
‘women’s clothes’ or kvenréttindi ‘women’s rights’, and while a kvenprestur is a
woman priest, a kvenlæknir is a gynecologist rather than a woman doctor (there
is no established feminine form of læknir ‘doctor’). At the same time, a woman
vicar, kvenprestur, could not be called prestakona, since that title was already
used for vicars’ wives when women began studying theology. In other words,
which element to choose may depend on already existing established com-
pounds with other meanings. It can also depend on the structure of the noun at
issue: If the male form ends with -maður ‘man, human’, it is natural to ex-
change this element with -kona ‘woman’. The masculine suffix -ari also seems
interchangeable with -kona, as in söngvari/söngkona ‘male singer/female singer’.
Nouns ending in -ur seem more likely to form compounds with kven-, like
kvenhöfundur ‘authoress’. However, masculine nouns to which kven- is prefixed
still remain grammatically masculine.
Another possible, but still very rarely used, way of forming feminine nouns
is to change the masculine endings -ur or -i to the feminine -a. One example is
the established form þula (f) ‘female announcer’ (on TV), derived from the
same base form as the masculine þulur (m) ‘announcer’. More recent examples
can be found in Vera, a magazine published by the feminist political party
Samtök um kvennalista ‘organization for a women’s list’. Since 1993 the editors
and the office manager are called ritstýrur (f.pl) and skrifstofustýra (f.sg),
respectively. These two words are new compounds with a feminine form of
-stjóri (m) ‘somebody governing something’ in the established words ritstjóri
‘editor’ and skrifstofustjóri ‘office manager’. The masculine -i is changed to
feminine -a, which causes an umlaut, and the result is -stýra. This is in full
agreement with productive morpho-phonological rules of Icelandic and

analogous with two older words: skólastýra ‘headmistress’, derived from
skólastjóri ‘headmaster’, and bústýra ‘housekeeper’, which has the masculine
equivalent bústjóri. The two new compounds have not been established in other
contexts, and they are likely not to be widespread, since the two older feminine
compounds are seldom used nowadays. Skólastýra was considered to be
discriminating against women, and a bústýra would mainly be found in the
older rural culture.
The fact that women often bear masculine occupational titles is considered
normal, since gender is regarded as a strictly grammatical category. But this
does not seem particularly consistent, if we take a title like hjúkrunarkona
‘nurse’, lit. ‘nursing woman’ into consideration. When the nurses’ training
school was incorporated into the university and more men chose this training
course, the occupational title was changed to hjúkrunarfræðingur (m) ‘with
nursing education’. The new title is a compound with fræðingur (m), a word
which seldom appears on its own but has a function similar to a derivative
suffix, meaning ‘a person who is educated in a special field’, similar to English
-logist, -logian, -ist, etc. It is derived from fræði ‘-logy’, cf. guðfræði/guðfræðingur
‘theology/theologian’ (guð ‘god’), líffræði/líffræðingur ‘biology/biologist’ (líf
‘life’). Other words have recently undergone a similar development.4 Skúrin-
gakona ‘cleaning lady’ was judged as pejorative and changed to ræstitæknir (m)
‘sanitary technician’, and the feminine form fóstra ‘nursery school teacher’, lit.
‘fosterer, educator’ has been changed to the masculine gender leikskólakennari
lit. ‘nursery school teacher’ to give the education higher status. If gender were
strictly grammatical, there would be no hindrance for calling a man X-kona
‘X-woman’, since it is considered natural to call a woman X-maður ‘X-man’.
This practice can be interpreted as illustrating the rule of masculine as un-
marked in current Icelandic; at the same time it shows that referential gender
has a tendency to overrule grammatical gender in the area of personal reference.
The use of the generic masculine can be interpreted as making women
invisible. However, it is not language itself that is sexist, but language use. This
suggests that masculine/male is the norm in language as well as in society.
Rögnvaldsson makes the following observation:
Here there also appears a considerable sex inequality which it is correct to
point out, though it is of a social nature rather than linguistic. A word such as
skáld ‘poet’ is neuter, as the definite article shows, skáldið, still people have
found a reason for coining the word skáldkona lit. ‘poet-woman’, but nobody
finds skáldmaður lit. ‘poet-man’ or karlaskáld lit. ‘man-poet’ necessary. An
Icelandic 175
even more remarkable example is hetja ‘hero’; it is a feminine word, but all the
same people have found a reason to coin the compound kvenhetja lit. ‘woman-
hero’ when referring to a woman!
(Rögnvaldsson 1990: 63, my translation)
Language and society are not isolated phenomena – language is a product of

and reflects society. When society changes it often influences language, and it is
plausible to assume that language changes can have an influence on society.
Iceland has changed rapidly from an agricultural country at the beginning of
the century to a modern industrialized society especially dependent on the
fishing industry. Most changes have taken place in the short period from World
War II until today, and it seems that society has changed so rapidly that these
changes are not yet represented in the language.
On the lexical level the Icelandic language has undergone major changes:
The lexicon has increased considerably as new technologies, new professions,
the sciences, computerization etc. have required new words. But the fact that
women are becoming more active and visible in politics, business, education
and other areas of Icelandic society has not had a major impact on the lexicon.
A popular example of women taking part in public life is that during a period
in the 1980s the three most powerful positions in Iceland were held by women:
the President, the Speaker of Parliament and the chair of the Supreme Court.
These changes have led to few noticeable changes in language. For instance, one
could expect that women would demand female/feminine or neuter job titles or
protest against generic masculine pronouns, but this has not been the case.5
3. Language and gender in Iceland
Iceland is a comparatively small nation with a small university and relatively few
researchers. Most graduate students need to go abroad for their doctoral degree.
This has the positive effect that Icelandic scholars are well-informed about
international research and open to new theories and methods. However,
feminist theories and research are at this stage only partially integrated in
Icelandic academia. There are several feminist researchers and a tradition of
gender studies within some fields, particularly the social sciences, history and
literature, but there are no linguists involved in larger studies on language and
gender. The two main forums for gender research in Iceland are Kvennasögusafn
Íslands (the Icelandic library for women’s history), founded by the historian
Anna Sigurðardóttir, who has done groundbreaking research on women’s
history (e.g. Sigurðardóttir 1985), and Rannsóknastofa í kvennafræðum (Centre

for Women’s Studies) at the University of Iceland. For examples of women’s
studies in Iceland see Kress & Traustadóttir (1997).
There are a small number of Icelandic articles and student papers on
language and gender issues.6 The leading researchers in the departments of
linguistics and Icelandic at the University of Iceland are mainly specialized in
the fields of philology, generative grammar, lexicography and language cultiva-
tion, and therefore there has not been much scope for either sociolinguistics in
a broad sense (e.g. discourse or conversation analysis), or research on language
and gender. Recent exceptions are a study on relationships between gender and
attitudes in a case of dialect change (Hjartardóttir 1997 and 1998) and Hilmis-
dóttirs (2001) study on youth language.
In addition, many active Icelandic feminists consider language change
irrelevant to the struggle for women’s rights (Kirstin Didriksen, p.c.). This fact
is connected with the Icelandic tradition of puristic language planning, which
has solid grass-roots support (Vikør 1995: 209ff). There has been sporadic
discussion during the last decades about how to make women more visible in
the language, but there has been no strong support for language reforms. For a
time it was not allowed to make job advertisements gender-specific, but in
reality this meant that feminine forms were forbidden, since masculine forms
were used generically (Theódórsdóttir 1985: 28). There have been a few femi-
nine occupational titles, like skólastýra, but the tendency is that they are used
less and less, and replaced by masculine titles.7 The common opinion today,
even among feminists, is that linguistic gender is strictly grammatical and has
no correspondence with gender bias.
Hornscheidt (1998) discusses different contexts that are of importance in
feminist language change: the linguistic, the cultural, the feminist, and the
political context. Purist language attitudes are characteristic of the cultural
context in Iceland, part of the explanation of avoiding feminization lies there.
Another part of the explanation is linguistic; if the masculine is the most
unmarked gender, it is apprehended by language users as having a weaker male
bias than is the case in German or English. If compared to German, which has
a similar three-gender system, and Swedish, which may use gender-indefinite
forms, Icelandic has a different strategy of avoiding gender bias in language.
Where German may employ feminization, e.g., Zuhörerinnen (f.pl) und Zuhörer
(m.pl) ‘listeners’, and Swedish tends to neutralization, i.e. åhörare ‘listeners’
(common gender.pl)8 (Magnusson & Jobin 1997: 156), the main tendency in
Icelandic is masculinization: áheyrendur (m.pl).
Icelandic 177
However, a few tendencies toward feminist attitudes to linguistic change

may be observed. In Vera, the following development can be found: In the first
issue in 1982, the legally responsible publisher was called ábyrgðarmaður (m) lit.
‘responsibility-man’; later this was changed to the non-personal noun ábyrgð (f)
‘responsibility’, then back to ábyrgðarmaður, and since 1994 the title is ábyrgð-
arkona (f) lit. ‘responsibility-woman’. In the same way designations for the
other staff members have changed from starfsmaður (m) lit. ‘work-man’ to
starfskona (f) lit. ‘work-woman’. Since 1993 the titles ritstýra (f) ‘female editor’
and skrifstofustýra (f) ‘female office manager’, which have been commented
upon earlier, are also used.
On the other hand, this development of feminine word forms does not have
a parallel in the yearbook of the Icelandic women’s rights organization. The
yearbook, called 19. júní (Yearbook 1979–1997)9, has used the masculine titles
ritstjóri ‘editor’ and ábyrgðarmaður ‘responsible publisher’ in every issue since
the first in 1979. This variation in language use suggests that linguistic change
is far from being a crucial issue in the feminist debate in Iceland.
4. The use of Icelandic occupational titles: A pilot study
In the following, one issue of the daily newspaper Morgunblaðið (January 29,
1995) will be analyzed. This newspaper’s political profile is independent liberal,
and with a circulation of approximately 50,000 it is the largest in Iceland. The
material was chosen with the intention of identifying a language use that is
widely regarded as “correct” current Icelandic use.
“Generic” masculine nouns are used to refer to women in current Icelandic
media discourse. Occupational titles and other nouns with a similar function,
such as nationalities and words denoting people skilled at or engaged in some
activity, like bílstjóri ‘driver’ or lesandi ‘reader’, are analyzed.
Occupational titles and nouns with a similar function were excerpted from
Morgunblaðið January 29, 1995, parts A, B and C, excluding cinema advertise-
ments and TV and radio programs. Table 2 shows how the words are distribut-
ed in subcategories according to gender and, in some cases, compound ele-
ments. Numbers in the tables count all written instances, many lexemes appear
repeatedly and in different inflected forms, and each instance was counted.
Men with occupations were mentioned three times as often as women in
the issue in question (147 instances of females and 429 instances of males).
Didriksen (1986) made a similar observation in a study of Faroese media
Table 2.Occupational titles in Morgunblaðið January 29, 1995

Occupational titles and other personal nouns Female Male Generic or
referent referent unspecific
referent
Neuter occupational titles:

compounds with -fólk, e.g. starfsfólk (pl) ‘staff’ – – 37
compound with kven-: kvenskáld ‘poetess’ 1 – –
other neuter occupational titles 6 14 21
Total 79 7 14 58
Feminine occupational titles:

compounds with -kona, e.g. leikkona ‘actress’ 9 – –
other feminine occupational titles, e.g. lögga ‘cop’ 21 12 18
Total 60 30 12 18
Masculine occupational titles:

compounds with -maður, e.g. starfsmaður ‘employee’ 16 93 280
compounds with kven- – – –
other masculine occupational titles, e.g. bílstjóri 94 310 511
‘driver’
Total 1,304 110 403 791
Occupational titles total 1,443 147 429 867
discourse. The Faroese and Icelandic societies have much in common; also,
there are structural linguistic similarities.
Women are precluded and neglected both at the visual and linguistic level in
the newspapers. It is only the men’s world that is described and verbalized,
while the women’s world is only mentioned a few times.
(Didriksen 1986: 103, my translation)
In Table 2 there were 79 instances of neuter occupational titles, 60 instances of

feminines and 1,304 instances of masculines. There were 376 different masculine
titles to be found (in a total of 1,304 instances), 22 feminine and 30 neuter (in a
total of 60 and 79 instances, respectively). Among the feminines are seven
feminine epicenes (such as au pair, fyllibytta ‘drunkard’, hetja ‘hero’ and lögga
‘cop’), but the others are compounds with, for example, -kona. Most neuter
nouns are collective words, there are 37 compounds with fólk ‘people’, such as
starfsfólk ‘staff’, launafólk ‘employees’ and listafólk ‘artists’. Most neuter nouns
are not specified for referential gender, but there is one example of the word
kvenskáld ‘female poet’, which is similar to the word skáldkona, observed by
Icelandic 179
Rögnvaldsson. This is strong evidence that maleness is regarded as the unmarked

case, since even a neuter occupational title is regarded as socially male.
The group of compounds with maður10 is remarkably large within the
category of job titles. There are in total 389 examples of such compounds (16
referring to females, 93 referring to males and 280 unspecific). In those words
maður has the function of “a suffix meaning ‘a person in a particular function,
role etc.’” (Webster’s Desk Dictionary 1983: 677), similar to -man and -person in
English. Compare, for instance, verkamaður to workman, worker and talsmaður
to spokesman, spokesperson. According to monolingual and etymological
Icelandic dictionaries (Íslensk Orðabók 1988, Magnússon 1989) maður has at
least two different elementary denotations: ‘human being’ and ‘male adult
human being’. In compounds, the generic meaning of maður is regarded as the
correct meaning without any male connotations (cf. Íslensk Orðabók 1988:614f).
Svavarsdóttir comments on these different denotations and asserts that:
Thereby there is a semantic opposition between the words maður and kona,
which denote human beings of opposite sexes. On the other hand, maður
sometimes has a wider reference. In the sentence Maðurinn er spendýr ‘Man is
a mammal’ the word maður denotes ‘human, human being’ and is therefore
not in opposition to kona in this case; on the contrary, it includes this meaning.
(Svavarsdóttir 1993a: 47, my translation)
Maður has a central role in Icelandic word coinage, it is a component of such

different words as mannamót ‘meeting, gathering’, almenningur ‘general public’,
menntun ‘education’, ungmenni ‘youngster’, and sjómennska ‘seamanship,
fishing’. An interesting case is the masculine noun kvenmaður ‘woman’, which
is derived from maður in its generic meaning ‘human’. Instead of making a
distinction between maður and kona, it is possible to make the distinction
between karlmaður ‘male human’ and kvenmaður ‘female human’.
Many occupational titles were coined at a time when women had no access
to being educated11 or to professions or occupations, other than vinnukonur
‘maids’, bústýrur ‘housekeepers’ etc., and therefore it was natural that all titles
were masculine, like lögmaður ‘lawyer’ and sjómaður ‘sailor’. New occupational
titles are often coined by adding a personal suffix to a noun or verb. They do
not always contain maður, but all of them are still masculine, like hönnuður
‘designer’ from hanna ‘design’(v) or hönnun ‘design’(n), or tölvari ‘computer
expert’ from tölva ‘computer’, or lögfræðingur lit. ‘with law education, lawyer’
from lag ‘law’ and fræðingur ‘educated person’. Although there are many female
professionals today, most of them do not seem to consider it discriminating that
they have masculine titles. Example (6) is taken from an article about SÍB, i.e.
Samband Íslenskra Bankamanna ‘The Icelandic Bank Clerks Union’:
(6) Árið 1977 voru konur 64% af starfsmönnum (m.pl)
bankanna […] Á afmælisárinu 1995 eru konur um 75%
félagsmanna (m.pl) SÍB […] Í aðalstjórn SÍB sitja sjö
manns (m.sg). [5 menn og 2 konur]
‘In 1977 64% of bank employees (m.pl) were women […]
In 1995 women make up 75% of the members (m.pl) of SÍB […]
Seven people (m.sg) [5 men and 2 women] sit on the SÍB
governing committee.’
(Morgunblaðið January 29, 1995: A 19, my translation)
Women make up 75% of SÍB members, but they are referred to three times with
words containing maður, and female bank clerks call themselves bankamenn
and never bankakonur. Nevertheless, there is a slight tendency to coin female
alternatives. In my material I found one alþingiskona (‘woman MP’, from
alþingismaður ‘MP’), six leikkonur (‘actresses’, from leikari ‘actor’), and two
verkakonur (‘female workers’, as opposed to verkamaður ‘worker, workman’).
The fact that some words with kona do exist indicates that some people regard
maður as marked and male-specific. In other contexts feminine compounds
with -stýra and derivatives with -kven have been observed.
Typical examples of generic masculines other than occupational titles are to
be found in job advertisements, such as that in example (7), where the mascu-
line plural indefinite pronoun þeir is used generically, when the company
explicitly encourages both men and women to apply for the job. In addition, the
inflectional ending of the adjective háskólamenntuðum (m.sg.dat) ‘with an
academic degree’ agrees grammatically only with manni (m.sg.dat), but not
with konu (f.sg.dat). In such cases of gender conflict in compound subjects,
agreement with the attributive adjective is normally determined by the element
nearest to the subject, and in that respect the sentence is grammatically correct.
At the same time, it is an example of the generic and unmarked masculine, since
the masculine element of the compound is placed before the feminine element
and thereby governs the gender of the adjective.
(7) […] auglýsir eftir háskólamenntuðum (m.sg.) manni (m.sg)
/konu (f.sg) […] Þeir (m.pl), sem hafa áhuga á þessu
verkefni […]
‘[…] advertises for a man (m.sg)/woman (f.sg) with an academic degree
(m.sg) […] Those (m.pl) who are interested in this project […]’
(Morgunblaðið January 29, 1995: B 20, my translation)
<DEST "gun-n*">
Icelandic 181
5. Conclusion
In the analysis of Morgunblaðið January 29, 1995, masculine nouns and pro-
nouns constitute the vast majority of words denoting human beings. A plausible
interpretation of this is that the masculine functions as the most unmarked
gender when referring to human beings. Contemporary Icelandic grammarians,
Rögnvaldsson (1990) and Svavarsdóttir (1993a), propose that the neuter is the
most unmarked gender, but this study suggests that there may be a difference
in markedness between human and non-human referents.
Another conclusion of the analysis is that the linguistic unmarkedness of
the masculine reflects a society where masculine/male is considered the norm,
since males are more visible in the media discourse and occupational titles are
masculine. These results tally well with Didriksen’s (1986) analysis of Faroese
media discourse. There has been sporadic discussion in Iceland during the last
decades about how to make women more visible in the language, but there has
been no strong support for language reforms. Explanations for this may be
found in the cultural and linguistic context. The common opinion today, even
among feminists, is that linguistic gender is strictly grammatical and has no
correspondence with gender bias. Still, Icelandic word coinage with -kona and
the recent feminine neologism -stýra, and instances of referential gender
agreement overruling grammatical agreement (as in example 3), indicate that
future change is not inconceivable if a small group of language users no longer
consider the generic masculine as generic or unmarked.
Notes
* I would like to thank Deborah Cameron and Kristinn Jóhannesson for valuable commen-
tary on previous versions of this paper (cf. Grönberg 1995). Special thanks to Maria Bonner
for thorough reading and useful information. I would also like to thank Katrín Helga
Andrésdóttir, Guðrún Haraldsdóttir and Dagný Kristjánsdóttir, Icelandic language users, for
encouraging the realization of this study.
1. Icelandic has four invariant particles (sem, er, and the archaic að, eð) that introduce
relative clauses. (In current Icelandic, eð is only used as part of the conjunction hvort eð er
‘anyway’.) In traditional grammars (Einarsson 1945: 70), they are treated as relative
pronouns, but in later works they are regarded as particles (Kress 1982: 113) or conjunctions
(Íslensk orðabók 1988: 3, 819) with relative functions. In modern Icelandic grammar they are
always treated as conjunctions.
2. The past participle has two main meanings in Icelandic, a verbal meaning and an
adjectival meaning (Friðjónsson 1989: 52f). Past participles in the neuter singular form the
so-called supine, which has a verbal meaning, in construction with hafa ‘have’ and a few
other auxiliaries: Hún hefur lokað (n.sg.nom) hurðinni (f.sg.dat) ‘She has closed the door’.
After vera ‘be’ and verða ‘become’ the past principle often has an adjectival meaning. Then
it is inflected for gender and number: Hurðin (f.pl.nom) er lokuð (f.sg.nom) ‘The door is
closed’. The past participle also forms a resultative-situative verb meaning in construction
with vera ‘be’ and verða ‘become’ (Kress 1982: 152f). There is a semantic difference between
the supine Hún hefur farið (n.sg) ‘She has left; she has accomplished the act of leaving’ and
Hún (f.sg.nom) er farin (f.sg.nom) ‘She is gone; she is in the state of being somewhere else’.
3. Information from Maria Bonner.
4. Compare with example (3). It implies that even though the grammatical gender of the
Icelandic word for ‘nurse’ is masculine, it may still be regarded as socially female, since there
are instances of feminine pronominalization in certain syntactic positions.
5. There still are considerable differences between women and men in working life. For
example, in 1995, women in full-time employment in the public sector earned only 73% of
the income of men in full-time employment in the same sector (Gender equality in Iceland
1995: 42), and an article in Morgunblaðið (March 3, 1995) reported that women with higher
education had only 64% of the salaries of men in similar positions.
6. Cf. Hjartardóttir 1984, 1997 and 1998, Ingimarsdóttir et al. 1994, Hróarsdóttir 1994,
Kvaran 1985 and 1994/1995, Stefánsdóttir 1987, Svavarsdóttir 1993b and 1994, Theódórs-
dóttir 1985, Kress 2000, Hilmisdóttir 2001).
7. Another example is words for office employees: skrifstofustúlka (f) lit. ‘office-girl’ was used
until the 1960’s. Then skrifstofukona (f) lit. ‘office-woman’ was used for a brief period, but
today mostly ritari (m) lit. ‘writer’ is used (information from Maria Bonner).
8. Common gender (called “utrum” in Swedish) combines the historical masculine and
feminine. For further discussion, see Hornscheidt, vol. III.
9. The yearbook derives its name from the date of the Icelandic Women’s Rights Day. In
1915, women (and men with a low income) were given suffrage on June 19th.
10. Maður consists of the root mað- and the masc. sg. nom. ending -ur. Depending on
morpho-phonological and historical factors, the root changes into mönn-, menn- and mann-
in certain surroundings (sg. dat. mann-i, pl. nom. menn-Ø, pl. gen. mann-a). These forms
can be found in related words, such as mannamót ‘gathering, celebration’ and almenningur
‘general public’.
11. In 1886, Icelandic women were allowed by law to attend higher education under special
conditions. In 1897 the first woman in Iceland graduated from upper secondary school, see
Jónsson (1977: 25 and 48).
Icelandic 183
References
Bjarnadóttir, Sigríður. 1996. Könnun á félagslegri líðan skólabarna (mögulegum streitum og

streitueinkennum) [A study of the social health of schoolchildren (Possible stress and
symptoms of stress)]. Unpublished paper. Örebro: Skandinaviska Ledarhögskolan.
Didriksen, Kirstin. 1986. Færøerne: Sprog og køn [The Faroe Islands: Language and gender].
Århus: Husets Forlag.
Einarsson, Stefán. 1945. Icelandic. Baltimore: Hopkins.
Friðjónsson, Jón. 1978. A course in Modern Icelandic. Reykjavík: Tímaritið Skák.
Friðjónsson, Jón G. 1989. Samsettar myndir sagna [Compound verb forms]. Reykjavík:
Málvísindastofnun Háskóla Íslands.
Friðjónsson, Jón G. 1990/1991. “Beygingarsamræmi með samsettu frumlagi” [Agreement
with conjoined nominative subjects]. Íslenskt mál og almenn málfræði [Icelandic and
general linguistics] 12/13: 79–104.
Gender equality in Iceland: National report to the fourth United Nations World Conference on
Women in Beijing 1995. 1995. Reykjavík: Ministry for Foreign Affairs.
Greenberg, Joseph H. 1966. “Language universals.” In Current trends in linguistics, vol. 3, ed.
Thomas A. Sebeok. The Hague: Mouton, 61–112.
Grönberg, Anna Gunnarsdotter. 1995. “‘Konan er líka maður’ [Woman is also a man]: An
analysis of generic masculine in modern Icelandic.” In Language and gender. Case studies
from a Swedish seminar [MISS 11], eds. Per Holmberg & Kerstin Nordenstam. Göte-
borg: Institutionen för svenska språket, Göteborgs universitet, 79–97.
Guðmundsson, Valtýr. 1922. Islandsk grammatik [Icelandic grammar]. København: Hagerup.
Haugen, Einar. 1976. The Scandinavian languages. An introduction to their history. London:
Faber & Faber.
Hilmarsson, Sverrir & Christopher Sanders & John Tucker. 1989. Íslensk–ensk orðabók/
Concise Icelandic–English dictionary. Reykjavík: Iðunn.
Hilmisdóttir, Helga. 2001. “Partiklarna þúveist och skiluru – ett isländskt ungdomssamtal
under lupp” [The particles y’know and y’see – a conversation between young Icelanders].
In Språk, kön och kultur: Rapport från fjärde nordiska konferensen om språk och kön.
[Language, gender and culture: Report from the 5th Scandinavian conference on
language and gender]. Göteborg, 6–7 oktober 2000, eds. Kerstin Nordenstam & Kerstin
Norén. Göteborg: Institutionen för svenska språket, Göteborgs universitet, 124–132.
Hjartardóttir, Þóra Björk. 1984. Kvennamál og karlamál. Um málfarslegan kynjamun
[Women’s talk and men’s talk. On gender differences in language use]. Unpublished
paper. Linguistics Dept. Reykjavík: Málvísindastofnun Háskóla Íslands.
Hjartardóttir, Þóra Björk. 1997. “Framburður, viðhorf og kynferði” [Pronunciation,
attitudes and gender]. In Íslenskar kvennarannsóknir: Erindi flutt á ráðstefnu um íslenskar
kvennarannsóknir 1995 [Icelandic Women’s Studies: Lectures held at a conference on
Icelandic women’s studies 1995], eds. Helga Kress & Rannveig Traustadóttir. Reykjavík:
Háskóli Íslands, Rannsóknastofa um kvennarannsóknir, 67–76.
Hjartardóttir, Þóra Björk. 1998. “Variablen ‘a’ i islandsk i relation til køn og holdninger”
[The variable ‘a’ in Icelandic in relation to gender and attitudes]. In Sprog, køn – og
kommunikation: Rapport fra 3. nordiske konference om sprog og køn. København, 11. – 13.
oktober 1997 [Language, gender and communication: Report from the 3rd Scandinavian
conference on language and gender], eds. Inge Lise Pedersen & Jann Scheuer. Køben-
havn: Rietzel, 87–94.
Hornscheidt, Antje. 1998. “On the limits of language planning: A comparison of Swedish and
German language change.” In Sprog, køn – og kommunikation: Rapport fra 3. nordiske
konference om sprog og køn. København, 11. – 13. Oktober 1997 [Language, gender and
communication: Report from the 3rd Scandinavian conference on language and
gender], eds. Inge Lise Pedersen & Jann Scheuer. København: Rietzel, 95–112.
Hróarsdóttir, Þorbjörg, 1994. “Konur skrifa bréf: Setningafræðileg athugun” [Women
writing letters: A syntactic study]. Mímir. Blað félags stúdenta í íslenskum fræðum
[Journal of the Society of Students in Icelandic Philology] 41: 23–30.
Iceland 1986: Handbook published by the Central Bank of Iceland. 1987. eds. Jóhannes Nordal
& Valdimar Kristinsson. Reykjavík. Central Bank of Iceland.
Ingimarsdóttir, Álfheiður & Birna Lísa Jensdóttir & Hrafnhildur Þorgrímsdóttir & Þóra
Magnea Magnúsdóttir. 1994. “Hóra, hlussa, perri, proffi” [Pejorative words for women
and men]. Mímir. Blað félags stúdenta í íslenskum fræðum [Journal of the society of
students in Icelandic philology] 41: 16–22.
Íslensk orðtíðnibók [Frequency dictionary of Icelandic]. 1991. ed. Jörgen Pind. Reykjavík:
Orðabók Háskólans.
Jónsson, Gísli. 1977. Konur og kosningar: þættir úr sögu íslenskrar kvenréttindabaráttu
[Women and elections: Passages in the history of the fight for women’s rights in
Iceland]. Reykjavík: Bókaútgáfa Menningarsjóðs.
Jónsson, Snæbjörn. 1927. A primer of Modern Icelandic. London: Oxford University Press.
Kress, Bruno. 1982. Isländische Grammatik. München: Hueber.
Kress, Helga. 2000. “Gender and gossip in the sagas.” In Sett och hört: en vänskrift tillägnad
Kerstin Nordenstam på 65-årsdagen [Seen and heard: Festschrift for Kerstin Nordenstam
on her 65th birthday], eds. Anna Gunnarsdotter Grönberg et al. Göteborg: Institutionen
för svenska språket, Göteborgs universitet, 191–200.
Kress, Helga & Rannveig Traustadóttir, eds. 1997. Íslenskar kvennarannsóknir: erindi flutt á
ráðstefnu um íslenskar kvennarannsóknir 1995 [Icelandic Women’s Studies: Papers given
at a conference on Icelandic women’s studies 1995]. Reykjavík: Háskóli Íslands,
Rannsóknastofa um kvennarannsóknir.
Kvaran, Guðrún. 1985. “Íslensk samheiti um konur” [Icelandic synonyms for women]. In
Íslenskar kvennarannsóknir 29. ágúst – 1. sept. 1985 [Icelandic women’s studies].
Reykjavík: Háskóli Íslands, 15–21.
Kvaran, Guðrún. 1994/1995. “Orð af orði: Ambindrylla og puðrureddi. Um heiti karla og
kvenna” [Old Icelandic synonyms for men and women]. Íslenskt mál og almenn
málfræði [Icelandic and general linguistics] 16/17: 171–208.
Magnússon, Ásgeir Blöndal. 1989. Íslensk orðsifjabók [Etymological dictionary of Icelandic].
Reykjavík: Orðabók Háskólans.
Magnusson, Gunnar & Bettina Jobin. 1997. “Gender and sex in German and Swedish.” In
Norm, variation and change in language, ed. Nyfilologiska Sällskapet. Stockholm:
Almqvist & Wiksell, 149–174.
</TARGET "gun">
Icelandic 185
Rögnvaldsson, Eiríkur. 1990. Íslensk orðhlutafræði: Kennslukver handa nemendum á háskóla-

stigi [Icelandic morphology: Textbook for university students]. 4th ed. Reykjavík:
Málvísindastofnun Háskóla Íslands.
Sigmundsson, Svavar. 1996. Rit um íslenska málrækt [Works on Icelandic language cultiva-
tion]. Unpublished bibliography. Reykjavík.
Sigmundsson, Svavar. 1998. Textar í íslensku fyrir erlenda stúdenta [Icelandic texts for foreign
students]. 4th ed. Reykjavík: Málvísindastofnun Háskóla Íslands.
Sigurðardóttir, Anna. 1985. Vinna kvenna á Íslandi í 1100 ár [Women’s labor in Iceland
during 1100 years]. Reykjavík: Kvennasögusafn Íslands.
Sjón. 1994. Augu þín sáu mig [Thine eyes saw me]. Reykjavík: Mál og menning.
Stefánsdóttir, Steinunn, 1987. “Um mun á orðavali kvenna og karla” [On the difference
between women’s and men’s choice of words]. Mímir. Blað félags stúdenta í íslenskum
fræðum [Journal of the society of students in Icelandic philology] 37: 17–27.
Svavarsdóttir, Ásta. 1993a. [1987] Beygingakerfi nafnorða í nútímaíslensku [The inflectional
system of current Icelandic nouns]. Reykjavík: Málvísindastofnun Háskóla Íslands.
Svavarsdóttir, Ásta, 1993b. “Gleðikonur og gleðimenn” [Gleðikona (‘prostitute’, lit. ‘joyful
woman’) and gleðimaður (‘joyful man’)]. Orðaforði heyjaður Guðrúni Kvaran [Vocabu-
lary harvested for G. K.], 21. júlí 1993. Reykjavík. (Internal festschrift), 14–17.
Svavarsdóttir, Ásta, 1994. “Agalegt skúffelsi eða Hvernig er orðafari kvenna haldið til skila í
orðabókum?” [Awfully disappointing or How is women’s language use presented in
dictionaries?]. Gullastokkur færður Gunnlaugi Ingólfssyni fimmtugum [A toybox given to
G. I., aged 50], 4. desember 1994. Reykjavík. (Internal festschrift), 13–17.
Svavarsdóttir, Ásta & Margrét Jónsdóttir. 1998. Íslenska fyrir útlendinga: Kennslubók í
málfræði [Icelandic for foreigners. Textbook on grammar]. 2nd ed. Reykjavík: Málvísin-
dastofnun Háskóla Íslands.
Theódórsdóttir, Guðrún, 1985. Kyn og kynferði [Gender and sex]. BA thesis. Háskóli
Íslands, Reykjavík.
Vera: Tímarit um konur og kvenfrelsi [Magazine on women and women’s liberty]. Vols.
1982–1997. Reykjavík: Samtök um kvennalista.
Vikør, Lars S. 1995. The Nordic languages: Their status and interrelations. Oslo: Novus Press.
Webster’s desk dictionary of the English language. 1983. New York: Portland House.
Yearbook. 1979–1997. = 19. júní. Ársrit Kvenréttindafélags Íslands [Yearbook of the Icelandic
women’s rights organisation]. Vols. 1979–1997. Reykjavík.
Year Book. 2000. = The Europa World Year Book. Vol. I. London: Europa Publications.
<TARGET "mar" DOCINFO
AUTHOR "Gianna Marcato and Eva-Maria Thüne"
TITLE "Gender and female visibility in Italian"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
italian
Gender and female visibility in Italian
Gianna Marcato
University of Padua, Italy
Eva-Maria Thüne
University of Bologna, Italy
1. Introduction
2. Categories of gender in Italian
2.2 The structure of Italian human nouns
2.3 Derivation
2.4 Compounding
2.5 Agreement
2.6 Pronominalisation
3. The semantics of human nouns
3.1 Semantic loading
3.2 Semantic asymmetries
3.4 Proper names
4. The representation of women and men in selected text types
4.1 Grammars
4.2 Teaching materials
4.3 Newspapers
4.4 Literary texts
4.5 Proverbs
5. Female and male discourse
6. Language politics
6.1 Language use and language politics
6.2 Splitting
6.3 Formation of new agentives
7. Conclusion
Notes
References
188 Gianna Marcato and Eva-Maria Thüne
1. Introduction
Italian (Italiano) belongs to the eastern Romance languages of the Indo-

European language family (cf. Lockwood 1972). It is the national language of
Italy, and also an official language in Switzerland, i.e. in the Canton of Ticino.
Italian is also used in Corsica, Istria, Dalmatia, Monaco, Malta, Greece, and
Tunisia. Italian has some 63 million speakers of which ca. 40 million are native
speakers. Italophone communities outside Italy owe their origin to successive
waves of emigration over the last centuries: There are more than 6 million
speakers of Italian in the United States, Argentina and Australia, and also
numerically significant sections of the population in Belgium, Germany, Great
Britain and France.
The modern standard language derives from the dialect of Florence, and
owes much to the literary prestige of Dante, Boccaccio and Petrarch. The
political unification of Italy in 1861 laid the basis for the development of the
national language; the spoken language developed largely independently of the
written variety, manifesting a large number of dialects and regional varieties (cf.
Lepschy & Lepschy 1978).1 Although in the 20th century, especially after the
Second World War, an overall process of homogenisation – mostly due to the
influence of the mass-media – has been observable, the presence of a wide range
of spoken varieties is still a notable characteristic of modern Italian, and most
speakers betray their regional origin to some degree in their accent and/or use
of vocabulary. Dialectologists stress that contemporary Italian is still based on
the coexistence of and continuous exchange between the standard literary
language and systems of dialects (cf. Pellegrini 1980).
Italian, one of the modern-day descendants of Latin, shows, for obvious
reasons, that the lexis is deeply influenced by Latin. There are cases of direct
continuity, where the standard Latin word for a given concept persists until the
present day. Also, from the Middle Ages on, Latin words have entered Italian
indirectly as a result of borrowing from external sources, such as other Ro-
mance languages, or English and German. Of course, Ancient Greek is another
important source of lexical borrowing.
In morphology, Italian reflects the clear separation of verbal and nominal
inflection characteristic of Latin, the latter inflection also including pronouns,
determiners and adjectives. Nouns belong to one of two grammatical gender
classes (feminine and masculine); they inflect for number (singular and plural),
but not for case (cf. Vincent 1988: 289).
Italian 189
Italian is a pro-drop language; the rich inflectional morphology of Italian

verbs systematically signals person and number of the subject in all finite
paradigms. Personal pronouns are not used, except for emphasis. It follows that
gender has no direct morphological representation in the Italian verb, except in
participial forms, which may show gender agreement with the object (when this
is a clitic, for example) in active structures, and with the subject in ergative and
passive structures, even when this subject is not explicitly represented by a noun
or a gendered pronoun, e.g., sei arrivata (f.sg) ‘you (female) arrived’, sei arrivato
(m.sg) ‘you (male) arrived’, i prezzi sono aumentati (m.pl) ‘the prices have risen’
(ergative), and la mela è stata mangiata (f.sg) ‘the apple was eaten’ (passive).
An important and very common morphological process in Italian involves
the use of evaluative nominal suffixes indicating the speaker’s attitude to the
entity in question, e.g.: ragazza ‘girl’, ragazzina, ragazzetta, ragazzuccia ‘little
girl’, ragazzona ‘big girl’, ragazzaccia ‘nasty girl’, ragazzotta ‘sturdy girl’.
Gender marking in Italian occurs on nouns, definite or indefinite articles,
demonstratives, possessives, numerals and quantifiers. Most of these categories
show morphological variation for gender and number.
2. Categories of gender in Italian

Grammatical gender has specific linguistic functions, but at the same time it is
related to people’s perception and experience of reality. Language is therefore
not neutral, because it influences the symbolic systems of speakers. And it is
precisely grammatical gender which is the principal category through which the
difference between female and male may be represented symbolically: “Gender
is not merely a grammatical category that regulates purely mechanical matters
of agreement, but is, on the contrary, a semantic category which reveals a
profound symbolism within language […]” (Violi 1986: 41, our transl.). Violi
demonstrates that in descriptions of categories of gender in linguistic theory
there is a persistent reduction of this semantic dimension to primarily formal
aspects. An examination of the most authoritative and widely-used Italian
grammars, such as Dardano & Trifone (1985), Renzi (1988), Schwarze (1988), and
Serianni (1989), reveals that this tendency continues to this day (cf. Burr 1995).
The categorisation of nouns in Indo-European languages was primarily
based on the semantic opposition between animate and inanimate, which is of
more ancient origin than that between masculine (male) and feminine (female).
Inanimate nouns were represented by the neuter grammatical gender, while the
opposition between feminine and masculine appears to have been a subcategory
of the animate gender. In the Romance languages the neuter gender gradually
disappeared and the original tripartite system evolved into one with only two
genders, masculine and feminine. This opposition has subsequently been
maintained and extended to include inanimate nouns (cf. Meillet 1921). As a
consequence, the Italian gender system is characterised by this basic distinction
between masculine and feminine, which are commonly thought to represent
“natural” categories.
Within the class of inanimate entities in current Italian, gender is an
inflectional phenomenon and gender assignment is semantically arbitrary;
within the class of animate entities a further semantic distinction is made
between human and non-human entities.
In the majority of cases, non-human nouns behave like inanimates: la tigre
(f) ‘the tiger’, la volpe (f) ‘the fox’, il bisonte (m) ‘the buffalo’, il castoro (m) ‘the
beaver’; in these cases of epicene nouns, if it is necessary to specify whether the
animal is female or male, this is achieved by female/male modifiers: la tigre
maschio ‘the male tiger’, il bisonte femmina ‘the female buffalo’. There are,
however, many cases of so-called natural gender assignment, in which gram-
matical gender expresses whether the particular animal is female or male. In
these cases, distinctions may be made in exactly the same way as within the class
of human beings, i.e. through different lexical roots, as in la vacca/la mucca (f)
‘the cow’, il toro/il bue (m) ‘the bull/ox’, or by unambiguous morphological
marking, e.g., by the feminine suffix -a or the masculine suffix -o: la gatta/il
gatto ‘the cat’.
2.2 The structure of Italian human nouns

The system of human nouns allows for the differentiation of feminine or
masculine gender both through inflectional morphology of the nouns them-
selves or through the use of satellite elements such as determiners and adjectives.
Regarding morphological distinctions, three subtypes can be identified.
First, there is a closed class of nouns where referential gender is expressed by
different lexical roots, which have the corresponding grammatical gender and
may, in addition, be followed by gender-specific endings. Many kinship terms
and agentive nouns belong to this class:
Italian 191
Table 1.Human nouns with lexical gender

Female-feminine (sg/pl) Male-masculine (sg/pl)
nuora/nuore ‘daughter-in-law’ genero/generi ‘son-in-law’

madre/madri ‘mother’ padre/padri ‘father’
suora/suore ‘religious sister’ frate/frati ‘religious brother’
Secondly, there is a nominal class, sometimes called mobile gender nouns, where
feminine and masculine terms share the same lexical root, which is then speci-
fied by means of the suffixes -a or -o in the singular, and -e or -i in the plural,
according to whether the referent is female or male, singular or plural. The suffixes
are bifunctional in that they represent gender and number at the same time.
Table 2.“Mobile gender” nouns
figlia/figlie ‘daughter’ figlio/figli ‘son’

ragazza/ragazze ‘girl’ ragazzo/ragazzi ‘boy’
impiegata/impiegate ‘female employee’ impiegato/impiegati ‘male employee’
cuoca/cuoche ‘female cook’ cuoco/cuochi ‘male cook’
In a limited number of cases, the suffix for the masculine singular is -e:
Table 3.“Mobile gender” with the masculine suffix -e
signora/signore ‘lady’ signore/signori ‘gentleman’
Thirdly, in the innumerable cases in which the nominal root does not permit
gender-specification by the suffixes -a and -o/-e, satellite elements, e.g., deter-
miners or adjectives, have to be used in order to achieve the overt marking of
referential gender. This class contains nouns such as nipote/nipoti (sg/pl)
‘niece, nephew’ or ‘granddaughter, grandson’, or giudice/giudici (sg/pl) ‘judge’,
and nouns derived from present participial forms, such as insegnante/
insegnanti (sg/pl) ‘teacher’ (cf. Section 2.3). In these cases, the suffixes -e and
-i only have the function of signalling number. Again, many agentive nouns
can be found in this class.
2.3 Derivation
Personal nouns may be derived from verbs, nouns or adjectives. In many cases,
the adjective and the derived noun share the same word form. Thus, an
adjective like contadina/o ‘peasant-like’ can easily be changed into a noun: la
contadina/il contadino (f/m) ‘the peasant farmer’. If the origin of the agentive is
a verb, the basis for the personal noun is a gerund (la diplomanda/il diplo-
mando, f/m ‘pupil in last year of high school’), a present participle (la presi-
dente/il presidente, f/m ‘the president’), or a past participle (la delegata/il
delegato, f/m ‘the delegate’).
One type of nominal derivation which is particularly associated with
agentive nouns specifies gender with the feminine suffixes -a/-e (sg/pl) or the
masculine suffixes -o/-i (sg/pl). Pairs of such nouns share the same lexical root
and the agentive suffix (-ai-, -aiol-, -an-, etc.):
Table 4.Nouns with the gender suffixes -a and -o
Agentive suffix Feminine Masculine Translation
-aila giornalaia il giornalaio ‘the newsagent’

-aiol- la pizzaiola il pizzaiolo ‘the pizza chef ’
-an- l’ortolana l’ortolano ‘the greengrocer’
-ar- la panchinara il panchinaro ‘the reserve (player)’
-in- la postina il postino ‘the postwoman/
the postman’
A second type of derived agentives has the suffixes -a/-e (f/m.sg) or -e/-i
(f/m.pl), which are also bifunctional:
Table 5.Nouns with the gender suffixes -a and -e
Agentive suffix Feminine Masculine Translation
-ier- la consigliera il consigliere ‘the councillor’

-on- la terrona il terrone ‘the southerner’ (offensive)
-tor- la pastora il pastore ‘the shepherd’
-or- l’assessora l’assessore ‘the town councillor’
There are a number of suffixes that form common nouns which have the same
form for both genders. In these cases, referential gender is specified by external
modifiers, e.g., definite articles:
Italian 193
Table 6.Gender distinction by article only

Agentive suffix Article + noun
-iatra la/lo psichiatra ‘the psychiatrist’

-ista la/il farmacista ‘the pharmacist’2
-ante la/il bracciante ‘the farm worker’
One very particular type of derivation of personal nouns is by means of the

highly productive suffixes -trice/-tore, as in l’attrice/l’attore ‘the actress/the
actor’, l’ambasciatrice/l’ambasciatore ‘the ambassador’. This represents a special
case, because gender is already unambiguously expressed by the agentive suffix,
while the final endings only specify number.
Another female-specific suffix is -essa, as in terms such as la studentessa/lo
studente ‘the student’, la professoressa/il professore. The problematical nature of
this suffix was already noted by Sabatini (1986: 116f), in that -essa, more often
than other feminine suffixes bears derogatory connotations (cf. also Cortelazzo
1995). Hence the existence of alternative forms which follow the morphological
patterns described above (l’avvocata, la presidente, la studente), which are used
alongside the -essa forms. Indeed, -essa is the only case of a feminine suffix in
Italian which has no masculine equivalent; the corresponding masculine terms
belong to one of many derivational classes, for example, the classes ending in -o
(l’avvocato ‘the lawyer’) and -e (il conte ‘the count’). Historically, morphological
feminines ending in -essa referred to the wife of the person denoted by the
masculine lexical base (la baronessa ‘the wife of the baron’), or to the female
counterpart of the referent denoted by the masculine term (la papessa ‘the
female pope’, from il papa ‘the pope’).
2.4 Compounding
While derivation is the traditional morphological pattern for the formation of
agentives, gender-specification can also be achieved by two more recent types
of compounding which combine a personal feminine noun with a (masculine)
nominal modifier (cf. Serianni 1989: 120). Examples of the type la donna x are:
Table 7.Donna + occupational term
Compound Translation
la donna (f) magistrato (m) ‘the woman magistrate’

la donna (f) medico (m) ‘the woman doctor’
Here, the grammatical gender of the compound is feminine and controlled by the
noun donna ‘woman’. Examples of the type il x donna are grammatically mascu-
line; they combine the masculine agentive il x and the feminine modifier donna:
Table 8.Occupational term + donna
Compound Translation
il medico (m) donna (f) ‘the woman doctor’ (lit. the doctor woman)
il giudice (m) donna (f) ‘the woman judge’ (lit. the judge woman)
One interpretation of the difference between these two female-specific forms is

that donna in the form la donna x emphasises the female component, while the
form il x donna highlights the “professional” component.3 Sabatini (1986: 120)
recommended avoidance of either compound type, because both types involve
a masculine component and thus fail to take advantage of the morphological
possibilities of inherently female-specific derivations (la magistrata, la giudice).4
Despite Sabatini’s recommendations, more and more examples of the types il
medico donna/la donna medico (for the more colloquial la dottoressa) can be
found (cf. Stewart 1987). Significantly, there are no parallel male-specific
expressions in Italian, such as, for example, il medico uomo.
2.5 Agreement
The gender of the noun has morphological consequences for other constituents
of the sentence. In Italian, there must be agreement between the noun and
other elements of the nominal group, such as articles, adjectives, and deter-
miners. Agreement must also be established between the noun and predicative
adjectives, cf. (1) and (2).
(1) a. la ragazz-a alt-a
the.fem.sg girl-fem.sg tall-fem.sg
‘the tall girl’
b. il ragazz-o alt-o
the.masc.sg boy-masc.sg tall-masc.sg
‘the tall boy’
(2) a. La zi-a è bell-a.
the.fem.sg aunt-fem.sg is beautiful-fem.sg
‘The aunt is beautiful.’
Italian 195
b. Lo zi-o è bell-o.
the.masc.sg uncle-masc.sg is beautiful-masc.sg
‘The uncle is good-looking.’
Agreement with gender in predication is also obligatory when there is a

compound verb requiring the auxiliary essere ‘to be’ as in (3):
(3) a. Simona è andat-a a Pisa.
Simona (female) is gone-fem to Pisa
‘Simona has gone to Pisa.’
b. Gianni è venut-o a Firenze.
Gianni (male) is come-masc to Florence
‘Gianni has come to Florence.’
The gender and number of the noun also have morphological effects on any
modifier that may be present in the nominal group, for example, relative clauses
(introduced by il quale/la quale ‘who, whom’) or participial verb forms (present
or past). Example (4) contains singular forms, example (5) contains plural
forms. Note that (5b) is ambiguous with respect to referential gender.
(4) a. La bambin-a della quale ti hanno parlato
the.fem child-fem about.the.fem whom to.you have spoken
Lucia e Marina è stat-a dimess-a oggi
Lucia and Marina is been-fem discharged-fem today
dall’ospedale.
from.the.hospital
‘The girl Lucia and Marina spoke to you about was discharged from
hospital today.’
b. Il bambin-o del quale ti hanno
the.masc child-masc about.the.masc whom to.you have
parlato Lucia e Marina è stat-o dimess-o
spoken Lucia and Marina is been-masc discharged-masc
oggi dall’ospedale.
today from.the.hospital
‘The boy Lucia and Marina spoke to you about was discharged from
hospital today.’
(5) a. Le turist-e, essendo arrivat-e in.ritardo,
the.fem.pl tourist-fem.pl being arrived-fem.pl late
hanno perso il treno.
have missed the train
‘The (female) tourists, having arrived late, missed the train.’
b. I turist-i, essendo arrivat-i

the.masc.pl tourist-masc.pl being arrived-masc.pl
in.ritardo, hanno perso il treno.
late have missed the train
‘The (male) tourists, having arrived late, missed the train.’
‘The (male and female) tourists, having arrived late, missed the train.’
In the case of coordinate nouns, Marcantonio & Pretto (1988: 323) make an
interesting observation, stating that when groups of nouns with which the
modifiers must agree are predominantly feminine, but also include one or more
masculine terms, agreement with the feminine is possible in the case of inani-
mate entities, as in (6), but the masculine is required with human nouns (7), cf.
also the corresponding priority of the masculine in Romanian (cf. Maurice,
vol. I) and Spanish (cf. Nissen, this vol.):
(6) La ros-a, la viol-a, il tulipan-o,
the.fem rose-fem the.fem violet-fem the.masc tulip-masc
la primul-a e la mimos-a, tutt-e
the.fem primula-fem and the.fem mimosa-fem all-fem.pl
profumat-e.
sweet.smelling-fem.pl
‘The rose, the violet, the tulip, the primula and the mimosa, are all
sweet-smelling.’
(7) Simona, Franco, Francesca e Renza, sempre
Simona (f) Franco (m) Francesca (f) and Renza (f) always
simpatic-i.
pleasant-masc.pl
‘Simona, Franco, Francesca and Renza are always pleasant.’
Generally, with coordinate plural nouns of contrasting gender the agreement of

the adjective is typically masculine, as in (8):
(8) ragazz-e e ragazz-i svegl-i
girl-fem.pl and boy-masc.pl intelligent-masc.pl
‘intelligent girls and boys’
Example (8) would seem to confirm the rule according to which gender
agreement is controlled by the last noun of the series. Serianni (1989: 199)
mentions examples like una sedia (f.sg) e un tavolo (m.sg) rossi (m.pl) ‘a red
chair and a (red) table’ vs. i minerali (m.pl) e le sostanze (f.pl) ferrose (f.pl)
‘minerals and substances containing iron’. However, this “rule” seems to hold
Italian 197
only for inanimate entities, but not for human beings. When the masculine
noun precedes the feminine noun, agreement is excluded both with the
feminine (9a) and with the masculine (9b). In this case, the only correct
solution is to repeat the adjective with each noun (9c):
(9) a. *ragazz-i e ragazz-e svegli-e
boy-masc.pl and girl-fem.pl intelligent-fem.pl
b. ?ragazz-i e ragazz-e svegl-i
boy-masc.pl and girl-fem.pl intelligent-masc.pl
c. ragazz-e svegli-e e ragazz-i
girl-fem.pl intelligent-fem.pl and boy-masc.pl
svegl-i
intelligent-masc.pl
‘intelligent boys and girls’
2.6 Pronominalisation
Within the Italian pronominal system, only personal, demonstrative and
possessive pronouns (third person singular and plural) have gender-variable
forms. As regards personal pronouns, the situation is complicated by the fact
that there are two (incomplete) sets of pronominal forms, the free (stressed)
forms and the clitic (unstressed) forms. The latter are pronounced as if they
were unstressed affixes of the verb, representing the unmarked pronominal
realisations of direct and indirect objects, e.g., non lo vedo (lit. ‘not it I see’, ‘I
don’t see it/him’).
Free (tonic) pronouns occur in both subject and (direct and indirect) object
function, whereas clitics (atonic) are restricted to the latter function. The free
pronouns display gender-variable forms only in the third person singular: lei/lui
(f/m) ‘she, her/he, him’ and essa/esso (f/m) ‘it’. The latter pair, also occurring in
the plural (esse, f / essi, m ‘them’), is stylistically marked and syntactically
restricted to non-focal positions, which are found mainly in written texts, for
example in scientific, administrative or literary language.
As regards the clitic set, gender is marked explicitly and systematically only
in the third person singular and plural direct object forms: la/lo (singular), le/li
(plural), cf. (10):
(10) a. La vedo stasera.
3sg.fem.acc I.see this.evening
‘I see her this evening.’
b. Lo vedo stasera.
3sg.masc.acc I.see this.evening
‘I see him this evening.’
The same distinction may in principle be made for the singular indirect object
forms (le/gli), but in practice the masculine form is preferred even in cases of
female reference. In colloquial speech, gli may replace standard le (f.sg.dat) or
loro (m/f.pl.dat), cf. (11):
(11) a. Le devo telefonare.
3sg.fem.dat I.have to.call
‘I have to call her.’ (female-specific)
b. Gli devo telefonare.
3sg.masc.dat I.have to.call
‘I have to call him/her/them.’ (male-specific and generic)
The use of personal pronouns in subject function represents the marked case in
Italian, because the subject is normally marked on the verbal forms. There are,
however, contexts in which explicit pronouns must be used, for example, when
emphasis is placed on the referent to which the pronoun refers:
(12) Lei ha scritto il testo, non Pietro!
she has written the text not Pietro
‘She wrote the text, not Pietro!’
Possessive pronouns agree in gender and number with the related noun (the
“possessed” entity); consequently, they do not reflect the possessor’s gender, but
behave like other nominal modifiers, cf. (13):
(13) a. Gianni mi ha presentato
Gianni (male) 1sg.dat has presented
su-a mogli-e.
poss.3sg-fem.sg wife-fem.sg
‘Gianni presented his wife to me.’
b. Simona mi ha presentato
Simona (female) 1sg.dat has presented
su-o marit-o.
poss.3sg-masc.sg husband-masc.sg
‘Simona presented her husband to me.’
Generally, in gender-specific contexts the choice of gendered pronouns is

unambiguous. In gender-indefinite contexts, where masculine forms are the
Italian 199
traditional prescription, referential ambiguities may arise.

Again, as in the case of indefinite pronouns, the choice of masculine forms
requires further masculine agreement forms in other word classes:
(14) Non è venut-o nessun-o.
not is come-masc.sg nobody-masc.sg
‘No one came.’
Quantifiers are inflected like adjectives and can indicate referential gender:
(15) Non ho visto nessun-a dell-e tu-e
not I.have seen none-fem.sg of.the-fem.pl poss.2sg-fem.pl
amich-e.
friend-fem.pl
‘I haven’t seen any of your (female) friends.’
In adjectival function quantifiers relate to the head noun’s gender:

(16) a. Ciascun-a ragazz-a ha portato un libro.
each-fem girl-fem has brought a book
‘Each girl brought a book.’
b. I ragazz-i hanno portato un libro ciascun-o.
the.masc.pl boy-masc.pl have brought a book each-masc.sg
‘Each boy brought a book.’
3. The semantics of human nouns
3.1 Semantic loading

In Italian, semantic loading is found in pairs of lexical items and idiomatic
expressions that appear to be parallel from a denotational point of view.
Overtly, the two members of the pair are only differentiated with respect to
referential gender, but in addition the female-specific term generally carries a
semantic loading which derives from underlying gender-related cultural values
and expectations.
The following examples are taken from Riolo (1995) and Secco (1995); they
include expressions drawn from various Italian dialects:
dare in sposa ‘to give in marriage’: dari in the Sicilian dialect has the same
principal meaning as Standard Italian dare ‘to give’; in extended use dari can
have the meaning ‘to give a girl to someone for marriage’: ti dugnu a mè figghia
‘I’ll give you my daughter in marriage’, m’ata a dari a vostra figghia ‘you must
give me your daughter in marriage’. In Venetian, some expressions exist that
underline the drama of young women married off (to an ill-matched partner)
purely for the economic interests of the family: destrigare na puta ‘get rid of a
girl’, negar una fia, which literally means ‘to drown a daughter’, but the expres-
sion signifies ‘marry off a daughter badly’; romper el colo a una puta has the
same meaning, its literal meaning being ‘to break a young girl’s neck’.
partorire ‘to give birth’: figghiàri is used in Sicily to mean ‘to give birth to a baby
girl’, while the verb parturiri, a refined word which derives from Standard
Italian, denotes giving birth to a baby boy (cf. Riolo 1995: 378). In Venetian, this
differentiation was taken a step further: the expression la ga fato na sépa, which
literally means ‘she’s produced a cuttlefish’, is an insulting way of saying ‘she’s
given birth to a baby girl’.
puttana ‘prostitute’, buttana in Sicilian: there is no corresponding male term,
either in Sicilian or in other dialects, or in the national language. Yet, from a
purely lexical point of view, this would seem to go against the trend of what is
generally considered to be a common process in Standard Italian, that is, the
tendency to create a masculine term when men begin to assume a role (especial-
ly a professional one) that was once carried out by women only.5 Maschiaccio
‘bad man’, masculazzu ‘bad man’, if referring to a man, preserves and reinforces
in Sicilian, as it does in a great deal of Italian oral usage, all the positive associa-
tions of “masculinity”, while in the corresponding term fimminazza, in Stan-
dard Italian femminaccia or donnaccia ‘bad woman, prostitute’, the process of
semantic degradation introduced by the suffix is unmistakable.
oltraggiare: the verb, derived from the noun oltraggio, means ‘to offend outra-
geously and abusively’ (cf. Cortelazzo & Zolli 1979–1988). Etymologically, it
means simply ‘to go beyond’. In legal language, it has a more precise definition;
here, it denotes ‘to commit an offence to the honour or prestige of a public
official or a public sector worker’. But if the word refers to a woman, it is
confined to the sphere of “decency”, in other words, it bears a clear sexual
connotation (cf. Marcato 1995b: 259–260).
There is no shortage of cases where it is actually overt lexical symmetry which

transmits the cultural difference relating to gender. An example is the opposi-
tion, found in Sicilian, between figghiu masculu ‘male child’ and figghia femmina
‘female child’ (cf. Riolo 1995: 378). Strictly speaking, the adjectives in these
expressions are redundant, since the male-female distinction is already encoded
Italian 201
in the gendered nouns. The adjectives arguably serve to activate the usual
connotations, positive for masculu, negative for femmina. On the formal level,
the two expressions are perfectly symmetrical, but in fact a clear gender-bias is
created. This example suggests that, if one wishes to discover the bias built into
the language, it may not be sufficient to focus on overt lexical asymmetries.
3.2 Semantic asymmetries

Another usage which displays semantic asymmetries is the ambiguous use of
uomo/uomini ‘man/men’ to denote either males or humans irrespective of
gender. Significantly, there is a tendency for nouns denoting abstract concepts
such as fratellanza or fraternità ‘brotherliness, brotherhood’, as well as for
collectives such as fratelli ‘brothers, siblings’ to exploit male expressions, rather
than equivalent female terms such as sorellanza ‘sisterhood’ (a parallel term for
‘brotherliness’ is lacking), and sorelle ‘sisters’, which also means ‘religious
sisters’. At the same time, a tendency can be observed to expressly include
women in complex expressions such as madre e padre ‘mother and father’,
maternità e paternità ‘motherhood and fatherhood’. In order to find gender-
neutral personal nouns in Italian, it is necessary to resort to lexical forms such
as le persone (f) ‘the persons’, gli individui (m) ‘the individuals’ or gli esseri
umani (m) ‘the human beings’. Irrespective of grammatical gender, these terms
have a gender-neutral lexical meaning.

As has already been noted, Italian has quite a varied range of morphological and
lexical resources at its disposal for the specification of gender. But, in actual use,
the feminine forms tend to be eclipsed by the masculine ones: The unmarked
generic term is always masculine, and the feminine is frequently morphological-
ly derived, as in professore/professoressa.6 This constitutes traditional usage, and,
obviously, once such use is established, it acquires the value of a norm. Howev-
er, it represents only a partial realisation of the potential offered by the language
itself. The principle of the masculine generic is present at all levels of the
language, because singular and plural masculine nouns are used to represent
either exclusively male referents or both male and female ones; for example, i
cittadini (m) ‘citizens’ can signify both ‘male citizens’ and ‘male and female
citizens’. This is a case of what Alma Sabatini (1987: 25) has described as
“grammatical asymmetries”, i.e. the disparity in the treatment of men and
women in grammatical forms, as opposed to “semantic asymmetries”, which

relate to discursive differences and lexical usage.
Apart from asymmetries that have to do with the rich agreement system of
Italian, another important area in which asymmetries are evident is in the use
of indefinite and impersonal pronouns. The masculine form is prevalent in the
use of pronouns, not only when reference is made to mixed groups, but also in
cases where referential gender is indeterminate, as in, for example: Chi ha visto
(m) questo spettacolo si è accorto […] ‘Everyone who has seen (m) this show has
realized […]’. And it is very unusual when addressing an audience – though it
can occur when addressing an exclusively female audience – to use expressions
such as: Chi è interessata (f) al dibattito […] ‘Anyone who is interested (f) in the
debate […]’.
3.4 Proper names

Asymmetrical use also occurs in the sphere of proper names. While common
nouns have a classificatory function, proper names have an exclusively referen-
tial function in that they are used to designate specific individuals. Part of this
designation is reference to the gender of the person in question.
The descriptions of proper names in various grammars (e.g., Dardano &
Trifone 1985, Renzi 1988, Schwarze 1988, Serianni 1989) do not convey a
homogeneous picture. The only point on which these grammars agree concerns
the definite article, depending on whether a man’s or a woman’s surname is
involved: La Thatcher e Brandt (cf. Sabatini 1987: 110).
From the lexical count conducted by Burr (1997), it emerges that the use of
the surname without a preceding article is the norm in reference to men, while
in referring to women the article is almost always used. A possible interpreta-
tion of this is that at some underlying level even proper nouns (i.e. surnames)
tend to have a gender bias in that female names are felt to be the exception
which must therefore be marked.
Sabatini (1987) criticised another aspect of the usage of proper names, that
is, the use of Signora ‘Ms, Mrs’ as a way of specifying reference to women while
men may be accorded a professional title, e.g., ai lavori coordinati dalla Signora
Roubet partecipa anche il Professor Ceccaldi ‘Professor Ceccaldi is also participat-
ing in the work coordinated by Ms Roubet’ (Sabatini 1987: 111; cf. also Burr
1997). In the corresponding expression for a man, the address term Signor
would rather be omitted: ai lavori coordinati dal Roubet […].
Italian 203
4. The representation of women and men in selected text types
4.1 Grammars
Studies on this issue are as yet generally lacking. Tatjana von Bonkewitz (1995)
conducted research on the use of the masculine generic in Italian grammars
used in schools. Bonkewitz’s aim was to find out whether Sabatini’s Raccoman-
dazioni per un uso non-sessista della lingua italiana ‘Recommendations for the
non-sexist use of the Italian language’ (1986) had been taken into consideration
in the writing of grammars. The analysis therefore included grammars pub-
lished before and after the Raccomandazioni.7
The findings demonstrate that women are poorly represented, partly as a
result of the prevalence of masculine generics and partly due to the vast number
of nouns (74.4%) referring to men only (of a total of 13,895):
The majority of examples are oriented towards men, the authors exclusively
address male pupils, the games that are cited are for boys, and women, if
present at all, have a subordinate role. There is a lack of positive and up-to-
date models for girls, and there is a lack of feminine forms for professions.
(Bonkewitz 1995: 108, our translation).
4.2 Teaching materials

To date very few critical analyses of existing teaching materials have been
carried out.8 Samson (2000) conducted research about gender representation in
textbooks for children between three and five years of age (scuola materna,
‘kindergarden’) and for children attending the compulsory years of schooling
between six and ten years (scuola elementare). The textbooks were selected from
the area of the humanities. The analysis was divided into a linguistic and a
contextual part. The former was based on the recommendations of Sabatini
(1986), the latter was subdivided into six sections: themes, people, illustrations,
settings, authors/authorities and exercises. The results show that there is a
disparate representation of gender in Italian school textbooks. First of all, there
is a predominant use of linguistic elements that subsume women under the
supposedly generic masculine form; this is accompanied by a lack of prominent
female characters in the illustrations. Secondly, a distinctly androcentric view of
society is presented due to a lack of female-related topics that go beyond mere
stereotypes. A similar difference in gender presentation emerged from an
analysis of grammar exercises in the textbooks, where it was found that the vast
majority of sentence subjects were boys (2,090 in comparison to 841 girls).

There are also studies on gender-based differences in education, which have
been promoted since 1989 by the Gruppo di pedagogia della differenza sessuale
‘Group on gender difference in pedagogy’ of the University of Verona, edited by
the educationalist Anna Maria Piussi. The aim of the group is above all to
present reflections on, and experience of, teaching with attention to gender
difference. For the first time the importance of female relationships and the
essential role of mediation that female teachers (the majority of teachers in
Italy) can offer, has been recognised.
The concept of a mother-tongue, in the sense of the first language the child
acquires in a communicative environment, thus becomes central (cf. Muraro
1989, Zamboni 1998), as regards both the use of language in schools on the part
of female and male teachers and students, and its symbolic value in the search
for a language which more closely depicts women’s real experience of life.
These studies throw light on a specific approach in the Italian scene,
represented especially by the group of philosophers, pedagogues, and linguists
in the group Diotima at the University of Verona. Philosophers like Luisa
Muraro (1981) and Chiara Zamboni (1995) point out that it is not sufficient to
reform the language at a political level only, because this merely touches on one
single aspect of the representation of women in language.
4.3 Newspapers
The first systematic analysis of newspapers was carried out by Alma Sabatini
(1987). The study was based on an examination of newspapers and magazines
from 1984, focussing particularly on the wording of job advertisements. Various
forms of grammatical and semantic asymmetry were identified: For example, in
the political pages, the constant use of a grammatically marked structure
involving the use of a determiner before the surname (e.g., la Falcucci) was
notable; significantly, this was never used for men. Generally, what emerges
from Sabatini’s work is the contrast in the linguistic representation of women’s
participation in society on the one hand, and the actual political roles that in
recent years have increasingly been filled by women, on the other. One need
only consider the persistent use of masculine terms such as il sindaco ‘the
mayor’, even in cases where the office is held by a woman. Inevitably, the use of
such terms leads to a whole series of morphological and semantic distortions. It
also becomes clear from Sabatini’s analysis that the majority of articles concern-
ing women were relegated to less prominent positions in the newspapers
Italian 205
(internal news pages, arts pages). Finally, the designation of women as a

separate category observed by Sabatini (1987) is still frequent in both spoken
and written language; cf. the following example: Questi popoli […] si spostavano
con le donne, i vecchi e i bambini […] cercando regioni più fertili. ‘These peoples
[…] moved together with the women, the elderly and the children […]
searching for more fertile regions’ (Sabatini 1987: 109).
Burr (1995, 1997) conducted an empirical analysis based on a corpus of
Italian newspapers published in 1989, focussing on the occurrence and use of
agentives and proper nouns. Her results confirm the findings of Sabatini, but
with some interesting variations: The system of the Italian language is not an
inherently sexist system (Burr 1995: 154), because it has at its disposal the
morphological means and derivational procedures necessary to achieve a
balanced representation of all agentives. In actual fact, the extended use of the
masculine form applied to female agents amounts to less than 1% of references;
the feminine agentive appears in 85.9% (of a total of 15,385) of the cases where
reference is to women, while the use of masculine terms (7.37%) or agentives
with external modifiers like donna (6.73%) is relatively rare (cf. Burr 1995: 155).
Thus in the majority of cases the linguistic system allows for an appropriately
transparent term to be used. What is striking, however, is that the occurrence of
agentive nouns with female reference in the whole corpus represents only a
small fraction of the total: 5.75%. Burr concludes that it is usage – not the
language itself – which is sexist. Consequently Burr judges the situation to be
“even more serious than the one previously described by Alma Sabatini” (Burr
1995: 156), because although in real life the situation has improved noticeably
in recent decades, in journalistic texts women still do not make news, unless in
traditional roles.
Research undertaken by Burns (1996) was also aimed at establishing the
extent to which Sabatini’s recommendations had taken effect in Italian usage.
The study was based on a corpus of magazine articles compiled in 1996.
Following a further suggestion of Sabatini (1987), Burns extended her study to
include a quantitative survey of the presence of men and women in pictures.
The result of Burns’s research confirmed the original findings of the study
conducted by Sabatini (1987).
Today, in the light particularly of recent editions of the Zingarelli dictio-
nary (from the 1995 edition onwards), a more diversified situation regarding
reference to women and men can be observed. On the one hand, recent
dictionaries display greater care, providing feminine forms and recording
neologisms generated by the changed role of women in society. On the other
hand, Chiantera (1999) has pointed out that many of these neologisms
continue to present a stereotyped image of women, and the explanations
themselves given in dictionaries remain wedded to a predominantly andro-
centric view of the world.
The situation in Italian is thus contradictory (cf. Thüne 1995), because
alongside masculine generic forms there is also a growing sensitivity towards
the use of gender-specific agentives. However, it still remains true that in the
field of agentives, women and men continue to use masculine forms even
when referring to female subjects, especially in reference to prestigious
occupations (e.g., Elisabetta Rasy, critico letterario e scrittrice ‘Elisabetta Rasy,
literary critic (m) and writer (f)’, Il Tirreno, 6 June, 1998). Using masculine
forms alongside female-specific ones in these situations is often an indication
of a changing social situation, which naturally produces a certain lack of clarity
in the use of the language and the degree of instability that feminine agentives
continue to display.
4.4 Literary texts

The analysis of the use of feminine agentives in a literary corpus conducted by
Stefania Spina (1995) involved the examination of the LIZ – Letteratura Italiana
Zanichelli, which contains the fundamental texts of a large part of Italian
literature from the 13th to the 20th century (500 works by 113 authors from
Francesco d’Assisi to Gabriele D’Annunzio). Spina’s analysis adopted both
linguistic/stylistic and social parameters.
Spina found that the most productive feminine suffix was -trice (50% of a
total of 182), followed by -essa and -aia.9 The most frequent agentives corre-
spond to jobs linked to small-scale trading, such as bottegaia ‘female shop-
keeper’, erbolaia ‘female greengrocer’, fruttolaia ‘female greengrocer’. These are
followed by agentives from the sphere of domestic work, e.g., massaia ‘house-
wife’. Spina stressed particularly the fact that the agentives from the first group
refer mainly to jobs of a low social status or even to jobs regarded with disdain
(bettoliera, ostessa, tavernara all meaning ‘pub landlady’). It is also interesting to
note that in the 19th century there is a noticeable increase of terms referring to
women in the field of education, with feminine terms like professoressa ‘female
professor’ and studentessa ‘female student’ first occurring in the literature at this
time (Spina 1995: 137–139).
Italian 207
4.5 Proverbs
Riolo (1995) and Secco (1995) have studied regional corpora, from Sicily and
the Veneto respectively, and both have emphasised a pattern of semantically
encoded misogyny in idioms and proverbs, which also emerges in current
standard usage. What is most interesting is the observation that in the use of
normative traditional linguistic formulae regarding marriage, or, more general-
ly in male–female relationships, the privileged partner is always the male. This
can be illustrated by the following examples:
a. Expressions which justify male harassment and violence:10
L’omo è cacciaturi, la fimmina è gaddina e s’aggiucca.
‘Man is a hunter, women are hens, and they perch.’ (R)
b. Expressions which underline men’s dominant position in the family:
Ogni gaddu canta nta lu so munnizzaru.
‘Every cockerel (husband) crows (dominates) on his own dung-heap (in
his own house).’ (R)
c. Expressions which accord to wives a deficient character:
A relòio, caval e sposa ghe manca sempre qualcosa.
‘A watch, a horse and a wife always have something missing.’ (S)
d. Expressions which maintain that it is an evil for a woman to remain un-
married:
Maritu bbonu e maritu tintu, amara a ccu nunn’avi nì bbonu e nì tintu.
‘Good husband or bad, if a woman hasn’t got one, she’ll be sad.’ (R)
e. Expressions which maintain female unreliability and perilousness:
A donna non si può credere, etiam poi che è morta.
‘You can’t trust a woman, not even after she’s dead.’ (B)
f. Expressions which contain advice on characteristics to look for or to
avoid in choosing a woman:
La deve eser matrona in strada, modesta in cesa, massera in casa e mata in let.
‘She should be a matron in the street, modest in church, a housewife in
the house, and wild in bed.’ (S)
g. Expressions which stress maternity as the only socially valued role of wom-
en:
essiri in statu nteresànti ‘to be pregnant’ (R)
h. Expressions which suggest gender differences through asymmetrical
judgements:
Le chiacchiere son femmine e i fatti son maschi.
‘Women chatter and men act.’ (B)
Sometimes prejudice, continually marked by lexical asymmetries, is more

radically ingrained, and involves the use of terms for which there is no neutral
alternative in normal use. Thus, a supposedly neutral approach ends up
perpetuating traditional views on what it is to be male and female. Examples of
this process can be found in the metaphorical use of terms relating to the lexical
field of sexuality. Males are presented in positive terms, while females are
metaphorically linked not only to weakness, but also to a general sense of
negativity. Fari fimmina, which literally means ‘to give birth to a baby girl’, in
some Sicilian dialects signifies ‘not to succeed in any given activity/business, to
fail’. A fimminìnna (R), literally ‘in the way of women’, can signify ‘simply,
roughly’, ‘to do something clumsily, imprecisely’; cf. an expression such as fàrisi
i cunti a fimminìnna (R) ‘do the sums/accounts inaccurately’. Although these
expressions have no relation to the social reality of women, they undoubtedly
perpetuate a certain stereotypical view of the world, and in particular, have the
power to transmit a distorted view of femaleness. Significantly, parallel expres-
sions such as *a masculina ‘in the way of men’ do not exist, presumably because
male behaviour is considered normal and therefore need not be signalled
specifically through language.11
As far as male and female speech is concerned, both at the level of classical
studies and at the level of oral traditions, recommendations about “speaking”
versus that of “keeping silent” appear to differ strongly depending on gender.
The constant advice to a woman is to “keep silent”, although by definition she
is the guardian of the mother-tongue, she is the one for whom the danger of
speaking is most strongly emphasised. 12
In treatises of the 16th century, the reason for which women were encour-
aged to keep silent was that “a woman is half a man” (cf. Chiantera 1995: 330)
and was therefore unsuited to anything other than the courtesies that render
relationships polite and the graceful conversation in which nothing substantial
is communicated: “like Echo, who never begins to speak of her own initiative,
but who, when someone speaks, is always ready to respond” (cf. Speroni 1560).
Her role is that of being a faithful mirror of her husband, a chameleon who
assumes the colours of her husband.
In popular culture, proverbs tend to perpetuate the idea that women talk
too much, and that they irritate males. In both cases, what is clearly involved are
Italian 209
communicative strategies aimed at limiting the opportunities for women to

express themselves.
From the 1970s onwards, the theme of gender differences in ways of speak-
ing began to be studied in Italian. Attili (1977) suggested that female/male
conversation could be interpreted as a collection of linguistic acts capable of
reflecting the way in which social relationships between individuals are struc-
tured: Female language, uncertain, anxious and less logical, is contrasted with
male language, didactic, authoritative and evaluative. This question was
discussed in terms of cultural diversity (cf. Tannen 1989). Attili & Benigni
(1977, 1979) aimed to demonstrate the different hierarchy of aims and the
different pragmatic means exploited by the two communicative patterns. The
male-female linguistic relationship is seen in terms of a relationship between a
majority and a minority, and is interpreted as a communicative process that
tends to create marginality, insecurity, the need for rules, and leads to frustra-
tion (cf. Attili 1977, Attili & Benigni 1977, 1979). In the same studies, the
authors point out that the strategies of male discourse reveal men’s desire to
keep women, and also children, dependent on their authority.
More recent approaches to speaking strategies show greater reluctance to
make generalisations that cannot be fully sustained. An important contribution
relating to the use of attenuating forms by women is that of Bazzanella &
Fornara (1995), which is based on an extensive conversational corpus. It shows
that there is no substance to this view: Women’s conversation is in fact not
characterised by more than average indecision, frequent attenuations and an
absence of “power”. Indeed, what emerges from a more careful analysis of
variables such as context and conversational role is that women are generally
more deeply involved in conversation.
The research reported in De Marco (1995) rejects the idea that the use of
the diminutive is characteristic of female speech, demonstrating that it is also
used in male speech, and stressing that if anything it is more extensive contact
with children that accounts for the quantitively greater use of such terms. The
choice of the diminutive is thus no more than a strategic choice in a given
communicative situation.
6. Language politics
The major source for the development of gender-neutral language in Italy is the
Raccomandazioni per un uso non sessista della lingua italiana ‘Recommendations
for the non-sexist use of the Italian language’ for schools and educational
publishers, published by Alma Sabatini in 1986 and adopted by the Prime
Minister’s office. In 1987 Sabatini published another work, Il sessismo nella
lingua italiana ‘Sexism in the Italian language’. As yet, the Commissione nazio-
nale per la realizzazione delle pari opportunità tra uomo e donna ‘The National
Commission for Equal Opportunities’ (i.e., the organisation which was origi-
nally responsible for the publication of both works) has failed to develop a
coherent strategy for the solution of the linguistic issues involved.
There may be three reasons for the current lack of progress. One may
simply be that, although potentially the linguistic (or more specifically the
morphological) structure of Italian might allow for gender to be expressed
systematically throughout the language, the traditional norms of usage seem to
resist such a development. Secondly, it is possible that socio-cultural develop-
ments over the last thirty years have encouraged the adoption of Anglo-Ameri-
can models, which have then brought in their wake a certain preference for
solutions typical of that linguistic culture, such as the adoption of “unmarked”
masculine forms. A third, and perhaps not unimportant, reason is the sarcasm
with which, even in learned circles, Sabatini’s recommendations were received.
6.1 Language use and language politics

It is interesting to note the initiative launched by the daily newspaper Il Sole –
24 ore, which in 1994 published an article entitled La grammatica della chiarezza
trasforma moduli e documenti ‘Grammatical clarity transforms official forms
and documents’, subtitled Il Codice di stile del dipartimento della funzione
pubblica indica le regole per facilitare la comunicazione scritta con i cittadini ‘The
Manual of Style of the Department of Public Affairs sets out rules to improve
written communication with citizens’.
Citing a law dated 9th December 1977 (no. 903,1), the article stresses the
need to make an effort so that the ban on “any form of discrimination founded
on sex regarding access to jobs” is actually respected, even if such discrimina-
tion is activated in “an indirect way, through press advertising, or any other
form of advertising, which indicates as a job requirement being a member of
one or the other of the two sexes” (our transl.). Emphasis is placed on avoiding
the spread of negative stereotypes, especially through automatic use of the
masculine on the assumption that this is the unmarked gender, and thus applicable
to both males and females. Many departments in public administration have
Italian 211
distributed the text of these recommendations to their personnel, but there is

no information available about the effects.
In recent years, especially Italian universities have increasingly been
showing signs of directing their attention to the problem, organising seminars
and training courses orientated towards the issue, for example at the University
of Bologna. One example worthy of note is the inter-faculty seminar held at the
University of Padua, Studi delle donne, as part of the European Community’s
Socrates project. This was aimed at establishing the Rete Tematica Athena of
Women’s Studies, which in the academic year 1997–98 devoted a number of
sessions to female writing.
In the south, the Archivio delle donne ‘Women’s archive’ at one of the two
Neapolitan universities (Istituto Universitario Orientale) is worthy of mention.
A sizeable specialist library has been assembled, which contains the largest
collection of periodicals relating to Gender Studies available in the south.
In the north, particularly important and well established is the work of
Diotima, a community of female philosophers established at the University of
Verona in 1984.
In 1997, the National Council of Research (CNR) financed an inter-univer-
sity research project entitled Segni linguistici al femminile nella storia, nella
tradizione orale, nella letteratura ‘Feminine linguistic signs in history, in oral
traditions, and in literature’. This project, which is coordinated by Gianna
Marcato of the Linguistics Department of the University of Padua, includes
numerous participants from Italian and other European universities. It also has
didactic aims, and discusses in a concise, but at the same time thorough and critical
way, the different issues connected to the theme of Women and Language.
6.2 Splitting
Sabatini (1987) recommended the consistent use of splitting, i.e. the use of
feminine as well as masculine forms, in order to avoid the use of the so-called
masculine generics.
Examples of long splitting are le parlamentari e i parlamentari ‘the female
and the male MPs’ and le senatrici e i senatori ‘the female and male senators’.
Abbreviated splitting can be realised only on the level of the determiner, for
example, la/lo studente (f/m, singular) ‘the student’ and le/gli studenti (f/m,
plural) ‘the students’.
6.3 Formation of new agentives

The formation of new agentives reflects a general tendency to eliminate
connotations that are felt to be derogatory or socially discriminating. For
example, alongside the term spazzina/o (f/m) ‘street cleaner’, there is now the
job title operatrice ecologica (f)/operatore ecologico (m) ‘ecology worker’. Recent
formations of this type tend to encounter some degree of resistance in everyday
usage. Sabatini (1985: 67) already pointed to the fact that for certain jobs which
were traditionally done by women, but are now also held by men, there is a
greater willingness to accept masculine forms which make the gender reference
clear. One example is levatrice ‘midwife’; instead of creating the morphological
parallel *levatore, coining a new masculine term ostetrico was felt to be prefera-
ble, which in turn led to the formation of the feminine ostetrica.
7. Conclusion
The problem of gender-fair usage in Italian remains unsolved. From this point
of view, the position of Chiara Zamboni (1995) and the research group Diotima
is particulary interesting, for they insist that concentrating on linguistic signs is
not sufficient in itself, but has to be accompanied by social and political
changes. A similar view has been put forward by the Italian linguist Patrizia
Violi (1986). In the course of her discussion of alternatives to female linguistic
invisibility, Violi (1997) notes that the language system is characterised by a
high degree of both inertia and internal flexibility. Thus, the introduction (or
elimination) of a given term may fail to produce the desired results; instead, the
linguistic system will tend to restructure the new semantic field on the same
lines as the preceding one (cf. Violi 1997: 92).
It does not appear that the discussion which started back in 1986 has led to
extensive changes in language use. Indeed, there seem to be two conflicting
strategies in current Italian, female visibility versus neutralisation (cf. Hellinger
1990). As regards gender-specific linguistic usage and the overt representation
of the female element, it seems that the situation is actually deteriorating, not so
much because of resistance to the introduction of new terms intended to denote
new female occupations and activities, but rather because of the tendency to
give up those gender differences that were traditionally well-established in the
Italian language, masking the new female roles with linguistic signs of male
gender. This goes against both tradition and the historically established gram-
matical rules of Italian.
Italian 213
When Sabatini’s study appeared, there were many people who were
reluctant to regard it as an invitation to engage in serious reflection on the
problem. Instead, attention tended to focus on certain provocative proposals
in Sabatini’s work, which aroused endless discussions about whether these
were aesthetically acceptable or not. Problems of this type concerned the
proposed use of dottora ‘woman doctor’ and avvocata ‘woman lawyer’ rather
than dottoressa and avvocatessa, or of medica ‘female physician’ vs. medico
‘male physician’.
One must be aware that there is a difference between using a masculine
plural noun such as alunni ‘pupils’ so as to also include female pupils (alunne)
on the one hand and using a masculine noun as an outright substitute for a
feminine one, as in the case of direttore used to refer to a female director, on the
other. In the first case, the masculine plural can be understood in the traditional
way as including reference to the female element, while in the second case the
masculine noun simply masks the underlying female reality and obliterates it.
Thus many have been led to the rejection of certain morphologically
acceptable feminines in -e (such as assessore), on the grounds that in Italian
their gender denotation is not realised in the ending (which provides informa-
tion only about number), but in the morphology of the article and of any
modifiers and participles that happen to be present and agree with the noun.
Nor is it any more productive to insist on the forced adoption of complicat-
ed forms, such as extensive splitting, acceptance of which is clearly problemati-
cal. In fact, it is important to distinguish those cases in which language actually
obscures important distinctions from cases in which it simply “does not mark”
them explicitly.
Notes
1. On the historical development of Italian cf. Migliorini & Griffith (1984), Serianni &
Trifone (1994); a standard dictionary is Ragazzini (1995).
2. The suffixes -iatra and -ista obviate the gender distinction only in the singular, for in the
plural they have distinct forms for the feminine (le psichiatre/le farmaciste) and the masculine
(gli psichiatri/i farmacisti).
3. The use of the lexicalised component in compound-agentives is limited almost exclusively
to women. Besides this form, there is also the pair ‘female/male’, which, however, is used
more frequently for non-human animate entities, for example, il gorilla femmina/maschio
‘the female/male gorilla’.
4. It is certainly quite remarkable that in cases where a feminine form actually exists, e.g.,
lavoratrice/lavoratrici ‘female worker/s’, the need is felt to resort to an even more strongly
marked form, i.e. lavoratrice donna ‘female woman worker’.
5. Consider, however, the admittedly jocular denomination of mammo for a ‘father who
assumes the role of mother’ cited by Cortelazzo (1995: 51).
6. Dardano & Trifone (1985: 133) write: “Such a preference is explained by the fact that the
masculine is closer than the feminine to being ‘neutral’ in value” (our transl.).
7. The following grammars were examined: Dardano & Trifone (1989), Della Casa (1988),
Pressini & Ravizza (1993), Sabatini (1980).
8. One exception are the reflections of Mirella Scriboni (1993) on the teaching of Italian as
a foreign language.
9. This once again raises the question of the non-stability of certain forms, in contrast to
others which are evidently more widely accepted (cf. Burr 1995).
10. The examples are taken from the cited texts: (B) indicates that the source is Bierbach
(1995), while (R) indicates Riolo (1995), and (S) Secco (1995).
11. Here we make reference to examples taken from Riolo (1995); these come from work
done on a specific corpus whose aim was to demonstrate asymmetries linked to gender in
Sicilian idiomatic lexis. However, a similar corpus based on a collection of the principal
dialect glossaries and the main historical dictionaries of the Italian language would undoubt-
edly confirm that these findings are valid for Italy as a whole.
12. This is clearly emphasised in Bierbach (1995), and for the 16th and 17th centuries, in
Chiantera (1995); it is also confirmed by Ceretta (1995: 360–363) for the end of the 19th and
the beginning of the 20th century.
References
Attili, Grazia. 1977. “Due modelli di conversazione.” Studi di grammatica Italiana 6: 191–206.
Attili, Grazia & Laura Benigni. 1977. “Retorica naturale e linguaggio femminile.” In Psico-
logica e retorica, eds. Guiseppe Mosconi & Valentina d’Urso. Bologna: Il Mulino, 85–91.
Attili, Grazia & Laura Benigni. 1979. “Interazione sociale, ruolo sessuale e comportamento
verbale: Lo stile retorico naturale del linguaggio femminile nell’interazione faccia a
faccia.” In Retorica e scienze del linguaggio, eds. Federico Albano Leoni & Maria Rosa
Pigliasco. Roma: Bulzoni, 261–280.
Bazzanella, Carla & Orsola Fornara. 1995. “Segnali discorsivi e linguaggio femminile: Evidenze
di un corpus.” In Donna e linguaggio, ed. Gianna Marcato. Padova: CLEUP, 73–86.
Berruto, Gaetano. 1980. La variabilità sociale della lingua. Torino: Loescher.
Biasini, Cristina. 1995. “Differenza sessuale e linguaggio: L’esempio del vocabolario sessua-
le.” In Donna e linguaggio, ed. Gianna Marcato. Padova: CLEUP, 63–71.
Bierbach, Christine. 1995. “Normes et représentations de comportement langagier: La parole
féminine dans les proverbes.” In Donna e linguaggio, ed. Gianna Marcato. Padova:
CLEUP, 267–284.
Italian 215
Bonkewitz, Tatjana von. 1995. “Lingua, genere e sesso: Sessismo nella grammaticografia e in
libri scolastici della lingua italiana.” In Donna e linguaggio, ed. Gianna Marcato. Padova:
CLEUP, 99–110.
Burns, Amanda M. 1996. Il sessismo nella lingua delle riviste settimanali. MA thesis, Universi-
ty of Melbourne, Australia.
Burr, Elisabeth. 1994. Corpus of Italian newspapers. Oxford: Oxford Text Archive.
Burr, Elisabeth. 1995. “Agentivi e sessi in un corpus di giornali italiani.” In Donna e linguag-
gio, ed. Gianna Marcato. Padova: CLEUP, 141–158.
Burr, Elisabeth. 1997. “Neutral oder stereotyp. Referenz auf Frauen und Männer in der
italienischen Tagespresse.” In Sprache und Geschlecht in der Romania. Romanistisches
Kolloquium X, eds. Wolfgang Dahmen & Günter Holtus & Johannes Kramer & Michael
Metzeltin & Wolfgang Schweickard & Otto Winkelmann. Tübingen: Narr, 133–179.
Cardona, G. Raimondo. 1976. Introduzione all’etnolinguistica. Bologna: il Mulino.
Ceretta, Maddalena. 1995. “L’immagine femminile negli articoli in dialetto di un setti-
manale cattolico vicentino.” In Donna e linguaggio, ed. Gianna Marcato. Padova:
CLEUP, 351–363.
Chiantera, Angela. 1995. “Le donne e il governo della lingua nei trattati di comportamento
cinque-seicenteschi.” In Donna e linguaggio, ed. Gianna Marcato. Padova: CLEUP,
329–339.
Chiantera, Angela. 1999. “Dire l’esperienza femminile con le parole nuove. Un excursus nei
dizionari novecenteschi.” In Dialetti oggi, ed. Gianna Marcato. Padova: Unipress, 275–288.
Cortelazzo, Manlio. 1995. “Perché non si vuole la presidentessa?” In Donna e linguaggio, ed.
Gianna Marcato. Padova: CLEUP, 49–52.
Cortelazzo, Manlio & Paolo Zolli. 1979–88. Dizionario etimologico della lingua italiana.
Bologna: Zanichelli.
Dardano, Maurizio & Pietro Trifone. 1985. La lingua italiana. Bologna: Zanichelli.
Dardano, Maurizio & Pietro Trifone. 1989. Grammatica italiana con nozioni di linguistica.
Bologna: Zanichelli.
De Marco, Anna. 1995. “L’influenza del sesso nell’uso dei diminutivi in italiano.” In Donna
e linguaggio, ed. Gianna Marcato. Padova: CLEUP, 87–98.
Della Casa, Maurizio. 1988. Scritto e parlato. Grammatica ed educazione linguistica. Brescia: La
Scuola.
Hellinger, Marlis. 1990. Kontrastive Feministische Linguistik. Mechanismen sprachlicher
Diskriminierung im Englischen und Deutschen. München: Hueber.
Lepschy, Anna Laura & Giulio Lepschy. 1978. The Italian language today. London: Hutchinson.
Lockwood, William B. 1972. A panorama of Indo-European languages. London: Hutchinson.
Marcantonio, Angela & Anna Maria Pretto. 1988. “Il nome.” In Grande grammatica italiana
di consultazione, ed. Lorenzo Renzi. Bologna: il Mulino, 315–332.
Marcato, Gianna. 1988. “Italienisch: Sprache und Geschlechter – Lingua e sesso.” In Lexikon
der Romanistischen Linguistik. Vol. 4, eds. Günter Holtus & Michael Metzeltin &
Christian Schmitt. Tübingen: Narr, 237–246.
Marcato, Gianna. 1995a. “Donna e linguaggio. Un rapporto difficile? Introduzione.” In
Donna e linguaggio, ed. Gianna Marcato. Padova: CLEUP, 21–48.
Marcato, Gianna. 1995b. “La gentildonna di Guascogna.” In Donna e linguaggio, ed. Gianna
Marcato. Padova: CLEUP, 249–265.
Marcato, Gianna, ed. 1998. Dialetti oggi. Padova: Unipress.

Meillet, Antoine. 1921. Linguistique historique et linguistique génerale. Paris: Champion.
Migliorini, Bruno & T. Gwynfor Griffith. 1984. The Italian language. London: Faber & Faber.
Mosconi, Giuseppe & Valentina d’Urso, eds. 1977. Psicologia e retorica. Bologna: il Mulino.
Muraro, Luisa. 1981. Maglia e uncinetto. Racconto linguistico-politico sulla inimicizia tra
metafora e metonimia. Milano: Feltrinelli.
Muraro, Luisa. 1989. L’ordine simbolico della madre. Roma: Editori Riuniti.
Pellegrini, Giovan Battista. 1980. Carta dei dialetti d’Italia. Pisa: Pacini.
Pressini, Laura & Gabriella Ravizza. 1993. Grammatica italiana. Milano: Mondadori.
Ragazzini, Guiseppe. 1995. Dizionario inglese-italiano/italiano-inglese. 3rd ed. Bologna:
Zanichelli.
Renzi, Lorenzo, ed. 1988. Grande grammatica italiana di consultazione. 3 vols. Vol. 1.
Bologna: il Mulino.
Renzi, Lorenzo & Giampaolo Salvi, eds. 1991. Grande grammatica italiana di consultazione.
3 vols. Vol. 2. Bologna: il Mulino.
Renzi, Lorenzo & Giampaolo Salvi & Anna Cardinaletti, eds. 1995. Grande grammatica
italiana di consultazione. 3 vols. Vol. 3. Bologna: il Mulino.
Riolo, Salvatore. 1995. “La misoginia nelle parlate siciliane.” In Donna e linguaggio, ed.
Gianna Marcato. Padova: CLEUP, 377–387.
Sabatini, Alma. 1985. “Occupational titles in Italian: Changing the sexist usage.” In Sprach-
wandel und feministische Sprachpolitik, ed. Marlis Hellinger. Opladen: Westdeutscher
Verlag, 64–75.
Sabatini, Alma. 1986. Raccomandazioni per un uso non sessista della lingua italiana. Roma:
Presidenza del Consiglio dei Ministri.
Sabatini, Alma. 1987. Il sessismo nella lingua italiana. Roma: Presidenza del Consiglio dei
Ministri.
Sabatini, Francesco. 1980. Lingua e linguaggi. Educazione linguistica e Italiano nelle scuole
medie. Torino: Loescher.
Samson, Amanda M. 2000. Buona e bravo – The representation of gender in Italian education.
Ph.D. dissertation, University of Melbourne, Australia.
Schwarze, Christoph. 1988. Grammatik der italienischen Sprache. Tübingen: Niemeyer.
Scriboni, Mirella. 1993. “Che sesso ha la grammatica? Alcune considerazioni sul ‘sessismo’
nell’insegnamento della lingua italiana.” Culturiana 5: 12–17.
Secco, Gianluigi. 1995. “Considerazioni sul ruolo della donna veneta dalla lettura critica di
alcuni proverbi popolari.” In Donna e linguaggio, ed. Gianna Marcato. Padova: CLEUP,
389–392.
Serianni, Luca. 1989. Grammatica italiana. Italiano comune e lingua letteraria. Suoni, forme,
costrutti. Torino: UTET.
Serianni, Luca & Pietro Trifone, eds. 1994. Storia della lingua italiana. Torino: Einaudi.
Speroni, Sperone. 1560. Dialoghi. Venezia.
Spina, Stefania. 1995. “Lessico, donne e mondo del lavoro: I nomi femminili di mestiere nella
letteratura italiana.” In Donna e linguaggio, ed. Gianna Marcato. Padova: CLEUP, 129–140.
Stewart, Dominic. 1987. “Forms for women in Italian.” The Italianist 7: 170–192.
Stoppelli, Pasquale & Eugenio Picchi, eds. 1995. LIZ 2.0: La letteratura italiana Zanichelli.
Cd-rom dei testi della letteratura italiana. Bologna: Zanichelli.
</TARGET "mar">
Italian 217
Tannen, Deborah. 1989. Talking voices. Repetition, dialogue and imagery in conversational
discourse. Cambridge: Cambridge University Press.
Thüne, Eva-Maria. 1995. “L’acqua in movimento. Questioni di genere grammaticale e lessico
femminile.” In Donna e linguaggio, ed. Gianna Marcato. Padova: CLEUP, 111–129.
Thüne, Eva-Maria, ed. 1998. All’inizio di tutto. La lingua materna. Torino: Rosenberg & Sellier.
Vincent, Nigel. 1988. “Italian.” In The Romance languages, eds. Martin Harris & Nigel
Vincent. London: Routledge, 279–313.
Violi, Patrizia. 1986. L’infinito singolare. Considerazioni sulla differenza sessuale nel linguaggio.
Verona: Essedue.
Violi, Patrizia. 1997. Significato ed esperienza. Milano: Bompiani.
Zamboni, Chiara. 1995. “Linguaggio e differenza sessuale: Il problema dell’arbitrarietà del
segno.” In Donna e linguaggio, ed. Gianna Marcato. Padova: CLEUP, 53–62.
Zamboni, Chiara. 1998. “Lingua materna tra limite e apertura infinita.” In All’inizio di tutto.
La lingua materna, ed. Eva-Maria Thüne. Torino: Rosenberg & Sellier, 113–134.
Zingarelli, Nicola, ed. 1998. Vocabolario della lingua italiana. 20th ed. Revised by Miro
Dogliotti & Luigi Rosiello. Bologna: Zanichelli.
<TARGET "bul" DOCINFO
AUTHOR "Tove Bull and Toril Swan"
TITLE "The representation of gender in Norwegian"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
norwegian
The representation of gender in Norwegian
Tove Bull
University of Tromsø, Norway
Toril Swan
University of Tromsø, Norway
1. Introduction
2. Historical and sociolinguistic background
3. The representation of gender in Norwegian
3.2 Agreement: Pronouns and adjectives
3.3 Lexical gender
4. Word-formation
4.1 Compounding
4.2 Derivation
5. Names and honorifics
6. Gendered idioms and frozen expressions
6.1 Expressions with mann
6.2 Expressions with kvinne/dame
6.3 Other asymmetrical expressions
7. Language change and feminist language planning
8. Conclusion
Notes
References
1. Introduction
Metaphorically, Norway has often been referred to as a laboratory of language

policy and planning. Many linguists consider the country particularly interesting
as an example of a modern national state where official language standardisation
220 Tove Bull and Toril Swan
in general and politically controlled language planning in particular have had a

great impact on the development of the written language from the late 19th
century onwards. As far as language and gender is concerned, however, Norway
is neither more, nor less of a laboratory than any other European country. The
question of language and gender did not – at least not in a significant way – enter
the agenda of any language planning agency or committee in Norway till the
1980s, long after it had started to evoke the interest of the general public. Why
there is such a discrepancy between general Norwegian language planning and
feminist language reform is worth discussing.
During the 1970s, the international development of sociolinguistics and the
accompanying interest in language and gender (gender being considered a
sociolinguistic variable) led to a reorientation within Norwegian linguistics.
Sociolinguistics was gradually being taught at university level, and young
feminist linguists addressed the topic of the relationship between language and
sexism (cf. Ryen 1976). Questions such as “Is the Norwegian language sexist?”
and “Is there a separate women’s language?” were raised and discussed among
the “new” feminists who got organised during the 1970s. The first person to
focus on questions like these, however, was neither a woman, nor a linguist.
Rolf Mikkel Blakar, a psychologist, published several books and papers on
topics of language, gender and power, and his work received a great deal of
attention during these first very active years (cf., e.g., Blakar 1973, 1975).
Out of this interest, two fields of study emerged: firstly, the linguistic and
communicative behaviour of men and women, and secondly, the linguistic
representation of gender in language. The latter area covers the study of
linguistic sexism and of non-sexist language planning. Although interrelated,
the two fields have nevertheless developed separately. Norwegian research in
these areas cannot be discussed or presented in isolation; it is part and parcel of
an international tradition.
As far as gender representation in language is concerned, research on
typologically more or less related languages shows many common features and
aspects, despite other interlingual differences. Firstly, the use of the masculine
gender for generic purposes seems to be universal. Thus, the noun or pronoun
referring to a man can also refer to human beings in generic contexts. This, of
course, may contribute to the idea that the generic or typical human being is
male. Consequently, it leads to the linguistic invisibility of women, by assuming
that male includes female, but not vice versa.
Secondly, almost everywhere the concept ‘male’ and/or the masculine form
is considered to be unmarked, while the concept ‘female’ or the feminine form
Norwegian 221
is seen as the marked exception. Thirdly, female terms are likely to be derived
from male ones, e.g., through suffixation, and finally, female terms and agent
nouns have a greater susceptibility to semantic derogation. All these aspects are
relevant to Norwegian.
The documentation of linguistic sexism led to calls for change, in Norway
as elsewhere. The Norwegian discussion as to what type of changes to advocate
had much in common with the solutions suggested in other language commu-
nities where linguistic sexism had been put on the agenda. As for the problem
of how to find appropriate ways to refer to females and males, individually and
generically, there was an option between two principally different solutions:
gender-neutralisation versus gender-specification. Gender-neutralisation aims
at eliminating differential treatment. Thus, any gender marking, grammatical
or lexical, should be eradicated or at least minimised in all generic contexts. In
practise, this means elimination of gender suffixes, to detract attention away
from the category of gender. The strategy of gender specification, on the other
hand, aims at making women linguistically visible by feminisation of human
agent nouns, through suffixation or lexically.
Norwegian women hold a leading political position compared to most
other western countries. The position of women is particularly noticeable in
national politics. Women have voted for 85 years: Norway was the second
nation in the world to introduce the vote for women in 1913. Some other
important dates with respect to women should be mentioned: the first professor
in 1912, the first minister (member of the King’s cabinet) in 1945, the first vicar
in 1961, the first bishop in 1993. To some extent, then, Norway has for a long
time had at least token women in most parts of the nation’s life. And as far as
the present is concerned, while there might be differences as to the extent of
optimism, no one would deny that Norwegian women have indeed come a long
way, and would seem to have an enviable position compared to women in many
other countries.
Thus while Norway is not entirely a country where women are truly
represented in every sphere, we would claim that this process is well under way.
If we compare the Norway of the 1950s and 1960s with today’s Norway, it is
obvious that not only are there fewer housewives (45% of the entire workforce
are women), there are also relatively large proportions of women in many high-
status professions, and almost entire equality of numbers in national politics.
Women are important in a way they certainly weren’t in the 1950s and 1960s.
2. Historical and sociolinguistic background
Norwegian is a Germanic language, belonging to the North Germanic (Scandi-

navian) branch, with approximately 4.5 million speakers. Since 1885 Norwegian
has had two officially recognised forms, nynorsk ‘New Norwegian’ and bokmål
lit. ‘book language’, both of which have developed into conservative and more
progressive or “radical” varieties in the course of the 20th century. Bokmål is the
written standard of the majority, while nynorsk is used by some 17% of the
population (cf. Vikør 1995). The names bokmål and nynorsk normally refer to
the written standards and not to any spoken varieties of Norwegian. There is
actually a third variety, essentially a very conservative variety of bokmål, called
riksmål ‘national language’; this is, however, not an official standard.1
Officially, there is no standardised form of spoken Norwegian. People are
free to speak their local dialects anywhere and in any situation or context.
Teachers are not allowed to correct or standardise the spoken language of
students in schools, nor would most of them probably want to. Dialects are used
in the media, the universities, and in the government administration. Thus, the
linguistic situation of Norway is rather different from that of the rest of Europe.
The reasons for the division into two written standards are rooted in the
earlier influence of Danish. Over a period of 400 years, Norway formed a so-
called personal union with Denmark and was under Danish rule until 1814.
Thus, bokmål was more or less directly derived from Danish, while nynorsk was
constructed as some sort of common denominator of the Norwegian dialects,
with an emphasis on the dialects in the western and central parts of southern
Norway. In a series of reforms, bokmål has gone through a process of Norwe-
gianisation, and similarly nynorsk has undergone a process of modernisation;
the result is that over the years the linguistic distance between the two stan-
dards has grown smaller. Today, the two standards are very similar, indeed,
and understanding one means understanding the other, even if there are
grammatical as well as lexical differences. Examples in the text are from both
varieties, or from dialects, but mostly from bokmål, the written variety most
extensively used.
3. The representation of gender in Norwegian
Like several of its Germanic cognates, Norwegian has inherited from Indo-
European a tripartite gender system (cf. Beekes 1995: 174). In present-day
Norwegian 223
Norwegian, the system of grammatical gender is fairly simple, and mainly

involves the definite and indefinite articles. More forms exist in the dialects
than in the standard languages, and more forms exist in nynorsk than in bokmål.
Norwegian also has natural, i.e. semantic gender, and there is some interaction
between these two types. Nouns may be marked for referential gender with
compounding elements or suffixes, the latter being mainly [+ female]. This type
of suffixation has been undergoing a rather drastic change over the past decades
in that the suffixes are disappearing.
There are, then, two types of gender-related marking: grammatical markers,
i.e. suffixed definite articles (or freestanding articles), and derivational suffixes
such as -ske, -inne (which, e.g., correspond to Engl. -ess).

The grammatical gender system of most dialects and the standard varieties
nynorsk and bokmål is basically tripartite: masculine, feminine, neuter. The
Bergen dialect as well as conservative bokmål and riksmål have a two-part
NP-system, with neuter and common gender, i.e. the collapse of masculine and
feminine into one common gender class. The common gender-marking article
is identical to the earlier masculine form (cf. Corbett 1991: 124). Norwegian
nouns neither have grammatical case inflections (except for the genitive marker
-s) nor true gender inflections (as was the case in Old Norse). Mostly, it is
impossible to tell what gender a noun belongs to from the form of the word
(Faarlund & Lie & Vannebo 1997: 150). A few nominal endings, however,
denote (grammatically) feminine or masculine words. Thus, words ending in
-er and -eri are always masculine and neuter respectively; nynorsk words in
-skap, -nad, and -leik are always masculine, while -ing is often feminine (cf.
Faarlund & Lie & Vannebo 1997: 96f).
Grammatical gender is a property of the noun, with agreement marked on
variable forms of articles and adjectives. The indefinite articles en (m), ei (f), et
(n) precede the noun and are freestanding. Norwegian has two sets of definite
articles, one of which is suffixed to the noun, and thus looks like an inflection.
Bokmål may have common gender -en versus neuter -et or a three-way system:
masculine -en, feminine -a, and neuter -et. Nynorsk patterns as the latter. The
other definite articles den (masculine/feminine) and det (neuter) precede the
noun phrase and are freestanding; however, the suffixed definite article is used
in a wider set of contexts than the freestanding one. Thus “feminine noun”
means that the noun must be used with a feminine article (or, that the word has
a feminine suffix/compound element). The gendered indefinite articles are

shown in (1a), and the definite articles in (1b,1c):
(1) a. en gutt (m) ‘a boy’
ei jente (f) ‘a girl’
et barn (n) ‘a child’
b. gutten (m) ‘the boy’
jenta (f) ‘the girl’
barnet (n) ‘the child’
c. den lille gutten (m) ‘the small boy’
den lille jenta/en (f) ‘the small girl’
det lille barnet (n) ‘the small child’
In the plural, gender is neutralised in bokmål (2a), but not in the dialects and
nynorsk (2b):
(2) a. (de) guttene ‘(the) boys’
(de) jentene ‘(the) girls’
(de) skipene/skipa ‘(the) ships’
b. (dei) gutane ‘(the) boys’
(dei) jentene ‘(the) girls’
(dei) skipa ‘(the) ships’
The grammatical gender of basic human nouns such as kvinne ‘woman’ or gutt
‘boy’, i.e. words referring directly to people, but not, for instance, to occupa-
tions, tends to correlate with their lexical gender, and frequently such personal
nouns have a generic term as well (which is often grammatically neuter):
(3) a. en mann (m) ‘a man’
ei kvinne (f) ‘a woman’
et menneske (n) ‘a human being’
b. en gutt (m) ‘a boy’
ei jente (f) ‘a girl’
et barn (n) ‘a child’
Every noun with a feminine suffix or basic nouns referring to women may be
used with the feminine article (but may also have the common gender article,
which is identical with the masculine): kone ‘wife’, dame ‘lady’, kjerring
‘wife/woman’, prinsesse ‘princess’, sangerinne ‘singer’. There are, however,
feminine nouns denoting animals or inanimates as well: kråke ‘crow’, ku ‘cow’, ugle
‘owl’, bok ‘book’, sol ‘sun’, kvise ‘pimple’, penn ‘pen’, dør ‘door’, hytte ‘cottage’.
Norwegian 225
Masculine articles are used with male-specific human nouns such as mann
‘man’ or bror ‘brother’. In addition, nouns referring to both women and men
(arving ‘heir’, professor ‘professor’, lege ‘doctor’, etc.) tend to be grammatically
masculine. As with the other gender classes, there are masculine nouns in all
semantic categories: baby ‘baby’, oter ‘otter’, kylling ‘chicken’, bil ‘car’, måne
‘moon’, storm ‘storm’, reklame ‘advertising/commercial’.
Neuter nouns mainly refer to inanimates: tre ‘tree’, regn ‘rain’, lys ‘light’,
solskinn ‘sunshine’, hus ‘house’, gress ‘grass’, vindu ‘window’, ark ‘sheet’, blad
‘leaf ’, but there are neuter nouns with personal reference as well, cf. menneske
‘human being’, barn ‘child’, nurk ‘baby, tot’, kvinnfolk ‘woman’, fruentimmer
‘female/woman’, mannfolk ‘male/man’, folk ‘people/nation’.
Masculine nouns have the highest frequency: More than half of all nouns
are masculine (cf. Beito 1986, Heggstad 1982), but there are no particular
semantic regularities involved: An inanimate noun such as måne ‘moon’ is
masculine, as is ‘storm’, while the Norwegian equivalent for Engl. ‘rain’ and
‘sunshine’ are neuter, and ‘sun’ is feminine. Nouns referring to animals are
mostly masculine, but may occasionally be neuter or feminine. To summarise,
all grammatical genders include nouns referring to persons, although the neuter
class comprises very few such nouns.
Grammatical and natural gender may interact, for instance in epicene
nouns (i.e. nouns with a fixed gender whose referent may be male or female);
most of these will be grammatically masculine, but may be used in conjunction
with feminine as well as masculine pronouns (cf. example 4):
(4) Professoren (m) kom inn i rommet. Hun/Han så ingen.
‘The professor came into the room. She/He saw nobody.’
The opposite situation will hardly ever occur, for the simple reason that there
are very few feminine epicenes. There are a couple of nouns, however, which
are invariably feminine in nynorsk, while in bokmål they may be feminine or
masculine, e.g., kjempe ‘giant’ and nattevakt ‘guard’. With these nouns, a
masculine personal pronoun would be used if they referred to males, and a
feminine pronoun if the referent was a woman.
3.2 Agreement: Pronouns and adjectives

Unlike English, but like German, Norwegian reflexive pronouns are not gendered:
(5) Hun/ han vask-et seg.

she/ he wash-ipf refl
‘She/he washed her/himself.’
Personal and possessive pronouns are gendered, but in the third person singular
only; there is no gender marking in first and second person pronouns or in the
plural, cf. Table 1:
Table 1a.Third person pronouns, personal and possessive in bokmål

Gender Personal Possessive
Singular masc han (ham) hans

fem hun (henne) hennes
neut det dets
masc/fem den dens
(inanimate)
Plural all genders de (dem) deres
Forms in brackets are oblique forms that differ from the nominative.
Table 1b.Third person pronouns, personal and possessive in nynorsk

Gender Personal Possessive
Singular masc han hans

fem ho (henne) hennar
neut det dets
masc/fem den (hans, hennar)
(inanimate)
Plural all genders de (dem) deira
Forms in brackets are oblique forms that differ from the nominative.
In addition to pronouns meaning ‘he’, ‘she’ and ‘it’, there is a third person
singular pronoun den, which is common gender, and used with reference to
non-human feminine as well as masculine nouns:
(6) a. Den boka (f)/vasen (m) der, gi meg den.
‘The book/vase over there, give it to me.’
b. Mannen (m)/dama (f)/barnet (n) er ikke her, *den er i stuen.
‘The man/lady/child, isn’t here, *it is in the living room.’
In bokmål, inanimates may not generally be referred to with han or hun – unlike
in nynorsk, and most dialects, where grammatical gender governs the choice of
Norwegian 227
pronouns; in these varieties, stolen (m) ‘the chair’ would be referred to as han
‘he’, while matta (f) ‘the rug’ would be ho ‘she’.
Pronominal reference to animals may be chosen either according to the
noun’s grammatical or to its referential gender, the latter being the normal
choice especially in the case of pets. The word hund ‘dog’ is masculine, but if the
animal referred to is female, she may be referred to as hun (or den):
(7) Hunden slikket hvalpene sine. Hun/Den var tydelig stolt.
‘The dog licked her puppies. She/It was obviously proud.’
As mentioned above, nouns referring to both men and women that are normal-
ly grammatically masculine, such as botaniker ‘botanist’, forfatter ‘author’, and
professor ‘professor’, may be referred to with either a masculine or feminine
pronoun (constructio ad sensum), if the gender of the referent is known, as in
(8a). Similarly, referential gender will also override grammatical gender in those
cases where the latter is neuter, cf. (8b):
(8) a. Tannlegen kom inn. Han/Hun sa at …
‘The dentist came in. He/She said that …’
b. Det kom et mannfolk (n)/ kvinnfolk (n) inn og han/hun sa at …
‘A man/woman came in and he/she said that …’
Traditionally, the masculine pronoun was used as the generic pronoun. Thus,
in most dialects, if the gender of the dentist in (8) was unknown, han would be
used as a generic form. In present-day Norwegian, feminine pronouns are now
possible in generic contexts, too (cf. Section 7).
Unlike languages such as English, Norwegian does not seem to have a
strong tendency to use the female/feminine pronouns for phenomena such as
storms, nations, or ships (cf. Romaine 1999). In Norwegian, ‘storm’, being a
masculine word, would be han if gendered at all. Ships, however, may be
referred to as ‘she’ (cf. Faarlund & Lie & Vannebo 1997: 326). In some dialects,
especially along the Norwegian coast, han ‘he’ is used in a number of contexts
where there is no referent at all, or no human referent; for instance, there are
weather statements with so-called empty han, as in (9), where hun ‘she’ would
be impossible:
(9) a. Han er fæl idag.
he is horrible today
‘The weather is awful today.’
b. Han blæs godt idag.

he blows well today
‘It is very windy today.’
c. Han er kald/varm idag.
he is cold/warm today
‘It’s cold/warm today.’
d. Vi står han av.
we stand him off
‘We’ll manage.’
Adjectives are normally not affected by the masculine–feminine gender distinc-

tion, but are marked grammatically for feminine/masculine vs. neuter gender
(as well as for number). Thus we find all genders represented in the singular as
in (10a–c) and the plural as in (11a–b). Plural adjectives show no gender
variation at all.
(10) a. jenta er snill/ei snill jente/den snille jenta (f)
‘the girl is good/a good girl/the good girl’
b. gutten er snill/en snill gutt/den snille gutten (m)
‘the boy is good/a good boy/the good boy’
c. barnet er snilt, et snilt barn, det snille barnet (n)
‘the child is good/a good child/the good child’
(11) a. snille jenter og gutter og barn
‘good girls and boys and children’
b. De er snille.
‘They are good.’
However, some particular adjectives may distinguish all genders in nynorsk,

radical bokmål and most dialects, e.g., liten ‘small’, as shown in (12):
(12) gutten er liten/jenta er lita/barnet er lite
‘the boy is small/the girl is small/the child is small’
3.3 Lexical gender

As is the case in most languages, Norwegian has a group of nouns that inherent-
ly denote women or men, and whose grammatical gender generally corresponds
to their referential gender, cf. (13):
Norwegian 229
(13) Female/Feminine Male/Masculine

kvinne ‘woman’ mann ‘man’
dame ‘woman, lady’ mann, herre ‘man, gentleman’
kjerring ‘(old) woman, wife’ kar ‘man’
pike, jente ‘girl’ gutt ‘boy’
Other inherently gendered nouns are kinship terms,2 words which denote
personal relationships, and a few miscellaneous terms:
(14) Female/Feminine Male/Masculine
mor ‘mother’ far ‘father’
søster ‘sister’ bror ‘brother’
tante ‘aunt’ onkel ‘uncle’
kusine ‘female cousin’ fetter ‘male cousin’
niese ‘niece’ nevø ‘nephew’
kone ‘wife’ mann ‘husband’
frue ‘wife, lady’ mann, herre ‘husband’
hustru ‘wife’ ektemann ‘husband’
brud ‘bride’ brudgom ‘bridegroom’
enke ‘widow’ enkemann ‘widower’
nonne ‘nun’ munk ‘monk’
hore ‘whore’ horebukk ‘whoremonger’
madonna ‘madonna’ [?]
dronning ‘queen’ konge ‘king’
Most of these nouns are in current use, though kjerring is rather derogatory
except in the meaning ‘wife’, which is widespread in the dialects (it is a deriva-
tion, now opaque, of the masculine noun karl ‘man’, which is the older form of
kar. Kvinne is probably the most frequently used general term for women,
though many speakers would prefer dame.
In Norwegian, like in many Indo-European languages, for instance German
and English, if a personal noun is derived, the female/feminine word will
normally be derived from a male/masculine form, not vice versa; the exceptions
are brudgom ‘bridegroom’ and enkemann ‘widower’, which are derived from the
feminine forms (cf. Bull 1985).
4. Word-formation
4.1 Compounding
In addition to lexical gender nouns, there are also a number of gendered
compounds, where the last element, normally a word from either group (13) or
(14), determines the referential gender denoted:
(15) a. karrierekvinne ‘career + woman’ = ‘careerwoman’
yrkeskvinne ‘occupation/profession + woman’
= ‘working woman’
sjøkvinne ‘sea + woman’ = ‘sailor, seawoman’
politikvinne ‘police + woman’ = ‘policewoman’
vaskekjerring ‘wash + wife/woman’ = ‘cleaning lady’
kontordame ‘office + lady’ = ‘clerk, typist’
stuepike ‘parlor + girl’ = ‘housemaid’
havfrue ‘ocean + lady’ = ‘mermaid’
prestefrue ‘vicar + wife’ = ‘vicar’s wife’
b. jordmor ‘earth + mother’ = ‘midwife’
barnehavetante ‘kindergarden + aunt’ = ‘nurseryschool
teacher’
helsesøster ‘health + sister’ = ‘public health nurse’
c. mormor/farmor ‘mother’s/father’s mother’
stemor ‘stepmother’
svigermor/datter/søster ‘mother/daughter/sister-in-law’
Several of the words in (15) are archaic today, e.g., kontordame, vaskekjerring,
due to changing social conditions or ideological pressure inspired by feminism,
and have been replaced by terms which do not express gender. Karrierekvinne
is new, replacing yrkeskvinne, as working outside of the home became the
norm, not the exception for women. All too frequently, however, karriere-
kvinne has acquired negative connotations, as had yrkeskvinne previously;
nowadays having a job is acceptable, but “having a career” is slightly more
suspicious, especially if the careerwoman in question has children. In (15b),
the nouns denote professions that used to be all female, but in which large
numbers of males are now working. Jordmor and helsesøster, however, remain
the official terms (newspapers might occasionally use the form jordfar, but
more or less humorously). Barnehavetante (and -onkel) is replaced officially by
førskolelærer ‘preschool teacher’, but may still be used infomally. In (15c) a
number of kinship compounds are listed.
Norwegian 231
4.2 Derivation
Gender suffixing in Norwegian characteristically has meant feminisation, i.e.
the suffixes are mainly used to derive feminine words from masculine ones.
However, gender suffixes are rapidly becoming extinct (cf. Swan 1992 and
1995). The masculine suffixes include -ør, -ist and -er, but the former two are
not productive at all. In addition, several -ør nouns are epicene, and always have
been: sjåfør ‘chauffeur’ – *sjåføse (but cf. massør ‘masseur’ – massøse ‘masseuse’,
etc.). The suffix -er in present-day Norwegian forms grammatically masculine
nouns, but is not generally regarded as a male suffix semantically (except by
very old people who may also use a corresponding female word ending in the
suffix -inne). The suffix -er derives agent nouns such as maler, from male ‘to
paint’, sykepleier ‘nurse’ from pleie syke ‘care (for the) sick’. The major feminine
suffixes are -inne and -ske, while -esse, -ine, and a few others may be considered
minor suffixes (cf. 16 and 17):
(16) Major suffixes denoting women: -inne and -ske
forfatter-inne ‘author’
venn-inne ‘friend’
lærer-inne ‘teacher’
sykepleier-ske ‘nurse’
forbryter-ske ‘criminal’
(17) Minor suffixes denoting women: -øse, -esse, -isse, -ine, -ise
mass-øse ‘masseuse’
baron-esse ‘baroness’
abbed-isse ‘prioress’
stud-ine ‘female student’
direktr-ise ‘directrice’
The above forms are the traditional ways of gender marking in Norwegian.3 As
to the grammatical gender of these suffixed forms, usage varies between
feminine and common gender in nynorsk. However, the suffixes are undergoing
a radical change in the language, i.e. not only are the feminine suffixes no
longer productive, they have in fact almost died out; their extinction undoubt-
edly being a result of the ideological process Norway started undergoing in the
late 1960s and early 1970s. As late as 1965, they were still going strong (cf. Swan
1992). The minor suffixes were never very productive in the first place.
Both -inne and -ske are borrowed from Middle Low German. They had very
different connotations, however, since -ske was almost entirely restricted to
negative or low status contexts, e.g., syer-ske ‘seamstress’, hykler-ske ‘female

hypocrite’. The suffix -inne was introduced in the Old Norse period. It was used
well into the middle of the 20th century for terms for women in many artistic
and intellectual professions or belonging to nationalities and ethnic groups, and
also to designate the wife of a high-status professional (bispinne, the wife of a
bisp ‘bishop’), and wives of various male members of the nobility and royalty
(hertuginne ‘duchess’, wife of hertug ‘duke’). The latter terms are in fact still
used today, and may also, of course, be titles not acquired by marriage but by
birth. More rarely, -inne was used to designate the female of certain animals
(ulvinne ‘she-wolf ’ from ulv ‘wolf ’); all of these are archaic as well. However,
even though most -inne nouns are more or less archaic today, there are some
present-day survivors, e.g., venninne ‘girlfriend’ (cf. venn ‘friend’), prestinne
‘priestess’ (cf. prest ‘priest’; the female word, however, is only used in the pagan
sense, while the masculine word is in current use to denote a minister or priest),
gudinne ‘goddess’ (cf. gud ‘god’), vertinne ‘hostess’ (cf. vert ‘host’), and perhaps
to some extent elskerinne ‘mistress’ (which with its masculine counterpart elsker
forms a symmetrical pair). For a complete account of such suffixes, cf. Swan
(1992 and 1995).
The disappearance of the female/feminine suffixes in Norwegian repre-
sents a case of relatively rapid language change which is clearly related to
changing social conditions; social and ideological pressures have indeed come
to bear on linguistic structures. Nevertheless, it is a curious change, since it is
only the suffixes that have come to be seen as trivialising and suspect, or even
derogatory; it is not gender marking per se, since compounding with -kvinne
is not suspect.
As far as words for men and women are concerned, official and feminist
policy requires a basic generic term and, if necessary, adjectival modification by
kvinnelig/mannlig ‘female/male’ or compounding with -kvinne/-mann. In fact,
the nominal system has been disturbed twice, first when people started entering
professions which had been dominated by either women or men – e.g., men
becoming nurses, and women becoming priests – and then due to ideological
pressure (cf. Section 7). Most terms for high-status or intellectual professions,
however, may refer to both women and men. The fact that they are grammati-
cally masculine is not significant with respect to pronouns, if the gender of the
referent is known. Apparently Norwegians do not usually consider grammatical
gender a salient feature – unlike the highly salient female/feminine suffixes –
and for many people masculine and feminine nouns are in one common gender
group. However, assumptions about the bearer of the title, if unknown, may
Norwegian 233
often be male-biased, e.g., in the high-status terms shown in (18), while titles
such as sykepleier ‘nurse’ (also grammatically masculine) will be female-biased.
(18) statsminister ‘prime minister’ biskop ‘bishop’
lege, doktor ‘physician’ biolog ‘biologist’
direktør ‘director’ politisjef ‘police superintendent’
dommer ‘judge’ advokat ‘lawyer’
politimester ‘Commissioner of Police, chief of police’
(also professor, pilot, journalist, president, ambassadør)
Like other Germanic languages, Norwegian has always had -mann words, such
as those in (19). Official policy has largely been to eradicate gendered forms,
including the -mann compounds. The ideal and extremely frequent pattern is
a basic generic form, and then – if necessary – gendered forms for specific uses.
Thus a stortingsrepresentant ‘member of Parliament’ may be a stortingskvinne or a
stortingsmann, a leder ‘chair’ (lit. ‘one who leads’) may be a formann or a forkvinne.
(19) Traditional usage Modern usage
tillitsmann lit. ‘confidence man’ tillitsvalgt ‘union
representative’
brannmann ‘fireman’ brannkonstabel ‘fire officer’
stortingsmann ‘parliament man’ stortingsrepresentant ‘parliament
member’
postmann ‘postman’ postbud ‘post carrier’
talsmann ‘spokesman’ talsperson ‘spokesperson’
formann ‘chairman’ leder ‘leader’
varamann ‘substitute man’ vara ‘substitute’
ombudsmann ‘ombudsman’ ombud ‘ombud’
Some of the -mann words, however, have been resistant to changes (cf. 20):
(20) sysselmann ‘(royally appointed) governor of Spitzbergen’
embedsmann ‘(royally appointed) higher civil servant’
fylkesmann ‘highest official of county’
tjenestemann ‘civil servant/state employee’
lagmann ‘judge (Superior Court)’
styrmann ‘first officer, co-pilot’
nordmann ‘(a) Norwegian’
rådmann ‘higher town official’
meglingsmann ‘arbitrator/mediator’ (in wage disputes)
Words like postmann, brannmann and a few more do not have corresponding
-kvinne forms, but have been entirely replaced with a gender-indefinite form
plus an optional adjectival modifier (kvinnelig/mannlig). On the whole then,
-mann words have died out, or are archaic/conservative. Once in a while,
however, one may find examples of general -mann words in the newspapers.
Thus a recent article quite unusually used ransmann ‘robber man’, riflemann
‘rifle man’ and gjerningsmann ‘perpetrator’ – as well as raner ‘robber’; all of
these referred to the same man!
Interestingly, there is no sign of people using the generic forms for women
only, as has frequently been the case in English (cf. Ehrlich & King 1994). On
national TV and in the newspapers, the man is a ‘spokesman’ and the woman
a ‘spokeswoman’, or they are both ‘spokespersons’. The case is similar for the
other frequently used words such as tillitsvalgt, leder, etc. Those persons who
still use the -mann words generally are old, or they do so as a political state-
ment, and indeed they claim to use them generically. The Norwegian Academy
of Science and Letters, for instance, uses the forms formann ‘leader, chair’ for
both women and men, which has become very unusual; it so happens that in
1999, the two formenn were women.
However, there are in fact some -mann words that are quite alive and well,
namely those listed in (20). These compounds will probably stay with us for a
long time, and relatively frequently one encounters feminine counterparts as
well (with -kvinne as a second element), and of course the form with premodi-
fying kvinnelig. The legal and official form, however, is the -mann form, so they
are the exception to the general trend and policy. The reason they are so
resistant to change is that almost all of them refer to high-status positions, and
in addition they are generally very old words, dating back to Old Norse times.
The word denoting ‘Norwegian’ is particularly difficult; the present authors
and quite a few other women dislike being called nordmenn, but there is no
alternative which seems acceptable.4 Similarly, landsmann ‘compatriot’ is problem-
atic, but seems to be disappearing. Generally, however, the system of Norwegian
personal nouns has become remarkably more generic or gender-neutral.
5. Names and honorifics
Naming practices and, not least, the honorific system, have been arenas where
gendering has been extremely dominant, and traditionally full of inequalities.
Norwegian has, however, changed a great deal in recent decades.
Norwegian 235
Norwegian first names are, with one or two exceptions, gendered, i.e. the
last element will clearly identify a male or a female referent (-geir, -grim, -er vs.
-vild, -hild, -a), or the entire name is defined as male or female (Tor, Bjørn vs.
Tove, Gunn). The two exceptions are boys’ names (Tore, Inge) which occasion-
ally are used for girls even though corresponding girls’ names exist (Tora, Inga).
With respect to last names the patriarchal tradition was strong until the
early 1970s.5 Women, on becoming wives, always took their husbands’ names,
as did their children. Per Ås, therefore, might marry Berit Li, who would become
Berit Ås, and their children might be Ole Ås, Nina Ås, etc. (occasionally Nina Li
Ås, i.e. with the mother’s maiden name as a middle name).
Since the 1960s, however, women have started to keep their own last names,
or at least add their own names as middle names; children of the marriage
might get their mother’s name. Indeed, since 1980, children automatically get
their mother’s last name unless the authorities are specifically informed that
they will use the father’s name. There are obvious class differences here, and
educated women will be more likely to keep their own names. A man may also
take his wife’s name, but this is rare. To some extent the type of name is also
important. If the man has a name ending in -sen ‘-son’, and the woman does
not, she is far less likely to change her name, since -sen names are more numer-
ous and therefore looked upon as more ordinary.
Traditionally, upper or middle class women could use their husbands’ titles,
a practice which is now archaic (though older people may retain it). The
formula would either be husband’s title + suffix + last name, or fru + husband’s
title + last name (cf. bispinne Ås ‘bishop-female Ås’; Fru doktor Ås ‘Mrs doctor
Ås’). Such usage was still prevalent as late as the 1960s.6
Some 40 years ago, when Norway’s first woman vicar was appointed in
1961, the papers were quite preoccupied with what title her husband would
have (Herr sogneprest? ‘Mr Vicar’? Herr prestemannen? ‘the vicar’s husband’?).
Conveniently, the bishop of Hamar, who is a woman, is unmarried.7
With respect to this usage, social rank as well as referential gender is
marked: The suffix -inne was used with high status wives (and is used to this day
with certain royals or nobles, e.g., hertuginne ‘duchess’). Also upper to middle
class is -frue (‘wife, married woman’; fru ‘Mrs’ is used as a honorific), and on a
low status level we have -kjerring or -kone (‘wife, married woman’), the latter
being more variable with respect to status (cf. 21). Prestekjerring and fiskerfrue
would thus be equally anomalous, but prestekone would be acceptable:
(21) prestefrue ‘minister’s/vicar’s wife’

fiskerkone/fiskerkjerring ‘fisherman’s wife’
kapteinsfrue ‘captain’s wife’
arbeiderkone/-kjerring ‘worker’s wife’
legefrue ‘doctor’s wife’
bondekone/bondekjerring ‘farmer’s wife’
With increasing numbers of women in relatively high positions, and changing

social values, such practices seem odd today. Nevertheless, it is quite clear that
women are far less likely to be addressed with their titles than are men.
Use of first names, however, is very common for both women and men. For
instance, Norwegian students address their professors by first name. In the
media, reference to politicians and other well-known persons will normally be
determined by the person’s likeability or whether she or he is well-known, not
by gender per se. Gro Harlem Brundtland, the previous Prime Minister, was
(and still is) frequently referred to as Gro, and increasingly fondly as time went
by and she became a veritable landsmoder ‘the country’s mother’ (derived form
of landsfader, i.e. a fatherly, old politician).
Both men and women are often referred to by last name in newspaper
articles, the full name being mentioned first, and thereafter only the last name.
Nevertheless, there used to be a much stronger tendency for men to be referred
to by last name and women by both (cf. Swan 1992).
The traditional honorific system included frøken ‘Miss’, fru ‘Mrs’, herr ‘Mr’,
but this is rarely used today, and if at all, only in business correspondence, in
which case Fr will be used for both frøken and fru, similar to Engl. Ms (cf.
Pauwels, vol. I). Previously, all these forms were routinely used, as well as
combinations with the husband’s title in the case of (middle class) women: Fru
professor Berg, Fru doktor Pettersen, but not *Fru bonde Haug ‘Mrs farmer
Haug’. Frøken and fru would originally be used only for middle class daughters
or wives. As late as 1911, even death announcements would be class differentiat-
ed in this respect (cf. Swan 1991). The loss of honorific titles, however, is not
simply, or even chiefly, a matter of feminist ideology; rather, it is a result of a
general democratisation process which also includes the loss of polite pronouns.
It should also be mentioned in this context that changing social practices
have led to various innovations in the lexical field of marriage and relationships.
The terms uekte barn/lausunge ‘ungenuine child’/‘loose child’ (‘illegitimate
child’) have disappeared, and a large percentage of children are in fact born to
parents who are not legally married, but often samboere ‘with-livers’, i.e.
‘partners’. Samboerskap ‘with-liver-ship’ (living together, cohabitation, cf.
Norwegian 237
ekteskap ‘marriage’) is an established term; the participants are called et sam-

boerpar ‘a with-liver couple’, or samboere (i.e. both man and woman call their
partner min samboer ‘my with-liver’). The degree of commitment and stability
varies, of course, but the term itself nevertheless confers a dignity wholly lacking
in earlier terms for relationships outside of marriage. Samboere may enter into
legal agreements designed to protect the parties involved (e.g., with respect to
inheritance), especially if there are children, and in some contexts have rights
and duties much like those in a marriage, but there is no ritual or ceremony.
Homosexual couples may enter into partnerskap ‘partnership’ (since 1993); this
involves a marriage ceremony in a judge’s office (but not, so far, a church ritual,
as the Norwegian state church is extremely homophobic), and the rights and
duties of the partners correspond exactly to those of heterosexual married
partners, for instance with respect to inheritance.
Other new terms which may be mentioned, very unpleasant ones, are
ekteskapsleir ‘marriage camp’, hentebrud (‘fetch + bride’) and, slightly older,
postordrebrud ‘mail order bride’. Ekteskapsleir is a phenomenon which is
entirely new, and a result of the recently opened borders between Norway and
Russia. It refers to a camp arranged for Russian women and Norwegian men,
providing opportunities for meeting potential partners; the women are (for
economic and other reasons, as they themselves say) very interested in finding
Norwegian husbands. Recently a women’s group pressed charges against the
manager of one of these marriage camps, claiming that he was a hallik ‘pimp’;
procuration, though not prostitution, is illegal. The Russian women’s response
consists of a counter charge; they feel, not unnaturally, that calling the camp
owner a pimp is tantamount to calling them prostitutes.
Hentebrud ‘fetch + bride’ and postordrebrud ‘mail order + bride’ are
derogatory terms for foreign women (often from Thailand or the Philippines)
who have married Norwegian men. Unfortunately, in some cases hentebrud is
an all too correct term, since in certain circles a variation on the theme of sex
holidays is to go abroad to get a wife, and has become very popular. These
women are described by the men as uncontaminated by feminism (and
therefore good wives, unlike the selfish and unfeminine Norwegian women);
unfortunately many end up abused and/or abandoned, without knowing any
Norwegian or having any friends. Divorce is often not an option, since they
may be returned to their home countries if they do leave their husbands.
6. Gendered idioms and frozen expressions
Languages are repositories of the linguistic practices of countless generations,

hence earlier social conditions may remain in the language long after they have
ceased to exist in society. Thus, like many languages, Norwegian displays a great
deal of asymmetry in the form and function of gendered phrases which reflect
centuries of inequality.
6.1 Expressions with mann

The original meaning of the word mann ‘man’ was probably ‘human being’ (cf.
also Latin homo; for an interesting account of the pairs homo/mulier and
vir/femina, see L’Hoir 1992). In present-day Norwegian menneske is used; this
word is neuter in Norwegian, while its Swedish cognate människa is feminine
(cf. also cognates in the other Germanic languages). It is in fact very difficult to
distinguish truly generic forms from masculine/male forms in various old
sayings and frozen expressions. Obviously, in many cases the male bias has been
reinforced because of social structures, even though the form itself may
originally have been generic. Thus, since mannskap ‘crew’ used to consist
entirely of men, the word had male connotations (grammatically, mannskap is
neuter). Members of the crew would be called sjømann/-menn ‘seaman/men’;
only recently has a corresponding word sjøkvinne ‘seawoman’ been created, and
is an approved word, though very rarely used.
Norwegian contains numerous lexical items and idiomatic expressions with
mann or its synonyms, while there are few with kvinne. In March 1999, the
Norwegian consumer ombudsman decided that papers may use the phrase rett
mann på rett plass ‘the right man in the right place’ in employment ads, since
this is an old saying. Ships are said to gå ned med mann og mus ‘sink with man
and mouse’ (sink with everyone on board), and if everyone went to attend an
arrangement one might say that de gikk mann av huse ‘they left man from
house’ (they all went). More depressingly, in the case of adverse events, people
may ta det som en mann ‘take it like a man’, but not ?som en kvinne ‘like a
woman’, or they may manne seg opp ‘to man oneself up’ (pull oneself together,
pluck up courage). A job or a task may kreve sin mann ‘demand its man’, i.e. be
a demanding job. When a man is mann for sin hatt ’man for his hat’, he is what
he promises to be. ‘Within living memory’ is called i manns minne ‘in man’s
memory’, and a generation may be called en mannsalder ‘a man’s age’. Then, we
have mannen i gata ‘the man in the street’ and menigmann, both meaning the
Norwegian 239
‘common man’. Finally there are, to mention but a few, manntall ‘census’ (lit.
‘man count’), mannevett ‘common sense’, altmuligmann ‘a jack of all trades’ (or
simply a janitor), and many more. En overmann is a superior, one who is better
at some skill, etc.; for many people, example (22) would not be anomalous at
all. The verb overmanne means ‘to overpower, overwhelm’; this is often a
physical overpowering by another person, but equally one may be overmannet
by one’s happiness, grief, or other feelings.
(22) Hun var hans overmann på alle felt.
‘She was better than him in all areas.’
A man may also be a levemann ‘man about town’ or a hedersmann ‘a man of

honour’. En viss mann ‘a certain man’, on the other hand, refers to the devil! A
woman would rarely be described as a hederskvinne (and it is not a word listed
in the dictionary) and never as a levekvinne (not even if she is mannbar ‘sexually
mature’); she might, however, be a luksuskvinne ‘luxury woman’, a concept for
which a male counterpart could hardly be found.
In addition to -mann formations, there are quite a few with herre/-herre
‘man, lord, gentleman’. Thus skapningens herre ‘the lord of creation’ may be
used to refer to a man or to people in general (as opposed to animals). A
person, perhaps even a woman, may leve som en herre ‘live like a lord’, cf. also
leve som en greve ‘live like a baron’. A ‘sumptuous meal’ is a herremåltid, and
naturally, this is what a herre og mester ‘lord and master’ deserves. Things shared
equitably are delt broderlig ‘shared brotherly’, but getting brorparten ‘the
brother’s share’ means getting most of it. Å være herre over means ‘to master’,
and herredømme is ‘command, mastery’, etc. A woman, too, might have
herredømme over sine handlinger ‘be in command over her deeds’ (be in control
– or in command over something else).
Most mann expressions, then, do not generally have kvinne or dame
counterparts, though some women, including the present authors, will deliber-
ately, but quite naturally, use terms such as i kvinnes minne ‘in woman’s
memory’, å være overkvinnet ‘to be overpowered’ and å kvinne seg opp ‘pull
oneself together’.
Some terms do have similar sounding opposites, but there is a difference in
connotations. Mannevett ‘man’s sense’ means common sense, which is a good
thing, whereas kvinnelist ‘female cunning’ is generally bad, and kvinnelogikk
‘woman’s logic’ is, of course, the opposite of “logic”.
6.2 Expressions with kvinne/dame

Searching for forms with kvinne and dame mainly yields forms like damesykkel
‘ladies’ bike’, which refer to physical objects made for women (and to which
there are corresponding herre-words). Otherwise there are few, but often
curious items, cf. for instance damebris and damevind ‘lady breeze/wind’ (light
breeze/wind). Worse, a jentekast ‘girl’s throw’ is a clumsy way of throwing
something. Strangely, the adjective or adverb jentete (derived from jente ‘girl’)
is not used about girls, corresponding to girlish/ly, but about men or boys being
effeminate or doing something clumsily. A famous Norwegian skier told the
media that he had skied jentete on a particular day, which caused quite a stir.
Norwegian also has the expression kvinne er kvinne verst ‘woman is woman’s
worst’ (women are other women’s worst enemy) – but also nød lærer naken
kvinne å spinne ‘need teaches naked woman to spin’ (corresponding to English
necessity is the mother of invention), which is colourful and rather attractive.
Kjerringa mot strømmen ‘woman against the current’ (a strongheaded woman)
is a character from a fairytale and has become the prototype of wilful women.
Traditionally, this expression was seen as bad, but is now also employed as a
term of pride by feminists. Kvinnelitteratur ‘women’s literature’, kvinne-
diskriminering ‘discrimination of women’, pikenavn ‘maiden name’, are terms
that are quite expressive, and which have no masculine counterparts as yet; the
former two are innovations in the language.
Finally, Norwegian has its share of words meaning ‘prostitute’ or ‘loose
woman’, with few corresponding male terms. This phenomenon is well-
documented universally (Schulz 1975). Examples from Norwegian are hore,
skjøge, ludder, gatepike, and tøs for the women, and the single horebukk ‘whore-
monger’ for men (though this is not really a true counterpart; cf. Bull 1990).
Most of these words are archaic; the word in use in present-day Norwegian is
prostituert, lit. ‘a prostituted one’. Note also the almost pleasant and comfort-
able sound of the female counterpart of hallik ‘pimp’ – horemamma ‘whore
mummy’, i.e. a brothel keeper or Madam.
All of the above suggests that Norwegian traditionally has been heavily
oriented in the direction of the male/masculine, except in negative, and
particularly sexually illicit, contexts. Mann, herre, etc., are used (pseudo-)
generically, or in positive contexts; kvinne, dame, etc., – to the extent that they
exist – denote something weak or negative, with few exceptions.
Norwegian 241
6.3 Other asymmetrical expressions

Social conditions may be the cause of asymmetrical lexical expressions, and
conversely, social changes may render the lexicon more symmetrical. One
example will illustrate how a recent law led to more linguistic symmetry. In
Norway, larger farmland properties (so-called odelsgårder ‘hereditary man-
ors/farms’) were until quite recently inherited by the eldest son, odelsgutten ‘heir
boy’, much like thrones are inherited by the eldest son (male primogeniture).8
Women could inherit, but only if there was no son at all; the daughter would
then be odelsjente ‘heir girl’. In 1971, the odelslov ‘male primogeniture law’ was
changed, and the previously asymmetrical pair odelsgutt – odelsjente became
symmetrical, referring to the firstborn male or female heir, respectively. Now a
girl may inherit the land or farm, even if she has younger brothers; it is age, not
gender, that determines the line of succession.
Social changes have brought about lexical changes in many areas. The form
ugift mor ‘unwed mother’ was used about women who had illegitimate children
(until a couple of decades ago, this was very shameful indeed!); needless to say,
there was no corresponding ugift far ‘unwed father’. In the 1970s, the form
alenemor ‘single mother’ (lit. ‘alone + mother’) was introduced, a term used for
both divorced and unmarried mothers, since the divorce rate was – and is –
climbing steadily. There were also cases of alenefar ‘single father’, though these
were quite rare. Nowadays, the masculine word aleneforsørger ‘single provider’
is the neutral, non-gendered form, with alenemor and alenefar as gendered
forms. These terms are used properly and symmetrically, but of course the
majority of aleneforsørgere are still women.
There is, on the whole, a strong tendency to provide both male and female
forms for terms like these; people tend to be much more aware of asymmetries.
Thus example (23), though perhaps trivial, illustrates this trend.
(23) Over femti festløver og -løvinner har samlet seg i befalsbarakken.
‘More than 50 male and female party animals have assembled in the
officers’ barracks.’
(Nordlys, 11 July 1998: 23)
Løve/løvinne ‘lion/lioness’ was used for male/female lions, respectively. The

word festløve, lit. ‘party lion’, means ‘party animal’, and used to be a frozen
expression; it is probably modeled on danseløve ‘dance lion’, i.e. a good dancer
or someone who is fond of dancing. The heading rendered in (24) is a femi-
nised form of an earlier generic term altmuligmann. As mentioned above, many
former and present -mann words can occur with -kvinne forms; embedskvinne
and tjenestekvinne, both meaning ‘civil servant’, are feminine forms of -mann
words which are the official, legal terms, and are found occasionally in newspa-
per language.
(24) Jazzens altmuligkvinne
‘the Jill of all trades of jazz’
(Nordlys, 11 July 1998: 19)
However, asymmetries still abound in numerous contexts. Thus while some

adjectives used about women, many of which are derived from nouns, may have
neutral or even positive connotations, namely kvinnelig ‘womanly’, pikeaktig
‘girlish’, and feminin ‘feminine’, others such as damete ‘ladylike’, kjerringaktig
‘(old)-wife-like’ and konete ‘wife-like’ are negative. In addition, a woman may
be mannhaftig ‘mannish, masculine’ (a word not used for men); she may be
jentete ‘girl-like’, however, this term is used more for men or boys. Many of the
above words may be used derogatively about men, denoting effeminacy,
weakness, etc. The corresponding male/masculine adjectives on the whole carry
positive connotations. Thus a man may be mannlig, mandig or maskulin, i.e.
‘manly’, a very positive description. Men or boys may also be mannfolkaktig
‘manlike’, or guttete/gutteaktig ‘boyish’, and again these will have positive
connotations. A girl may be called a guttejente ‘boy girl’ (tomboy). Finally, a
man may ta en mandig beslutning ‘make a manly decision’; a woman cannot
really do so, and neither man nor woman may *ta en kvinnelig beslutning ‘make
a womanly decision’.
Exclamatory phrases like For en kjerring! ‘What a wife, woman!’ may be
used about a man (a weak effeminate man) or a woman (a nasty, difficult
woman), but the corresponding For en kar! ‘What a man, fellow!’ would
normally be very positive for men, unless used ironically, and is not used about
women at all except in sentences such as Hun ter seg som en kar ‘She behaves
like a fellow’.
Other asymmetries include reference to royal couples. If the monarch is a
king, as in Norway at present, the king and the queen may be, and frequently
are, referred to collectively as kongeparet ‘the king couple’. In Denmark the
reigning monarch is a woman, a dronning ‘queen’; however, she and her
husband are never referred to in the Norwegian media as *dronningparet ‘the
queen couple’. Instead, the term regentparet ‘the regent couple’ is used, or – as
is normally the case in Denmark – dronningen og prinsen ‘the Queen and the
Prince’. Typically, the Danish Queen’s husband is not King of Denmark, while
Norwegian 243
the wife of the King of Norway is Queen of Norway. Legally and socially,
however, the Queen of Norway and the Queen of Denmark’s husband play the
same roles, namely as consorts, and without any constitutional rights and
duties. (Cf. also enkedronning ‘queen dowager’ but no corresponding *enke-
mannkonge ‘king dowager’.)
As mentioned above, phrases and compounds tend to have a male bias, and
if there are any feminine counterparts or near-counterparts, these are almost
invariably less positive. Men’s manhood (manndom) is mentioned and/or
celebrated in various terms which do not have feminine counterparts, thus
manndoms glans/prakt ‘the glory of manhood’ and manndomskraft ‘vigour of
manhood’, but no *kvinnedomskraft, and å tilbringe sitt manndomsliv ‘to spend
one’s manhood-life’, but no *kvinnedomsliv.
There are two sets of terms denoting ‘unmarried woman/man’, namely
gammeljomfru ‘old maid’ and ungkar ‘young + fellow’ (bachelor), which are not
symmetrical with respect to connotations, and also peppersvenn ‘pepper boy’
and peppermø ‘pepper maid’, used about individuals turning 30 without being
married. The members of the latter pair seem to be used symmetrically for
women and men; they are only used humorously in the present-day language,
since marriage is no longer a prerequisite for a woman any more than for a
man. The connotations of ungkar/gammeljomfru, however, are vastly different,
and indeed only ungkar is ever used in present-day Norwegian (and, occasional-
ly, an innovation ungkarskvinne ‘bachelor woman’).
Three terms for men and women deserve special mention. Gressenke ‘grass
widow’ and gressenkemann ‘grass widower’ in fact form a symmetrical pair,
although their use in this century had different motivations. Thus a woman
would normally become a gressenke because her husband’s work took him away
from their home; a man, however, traditionally was a gressenkemann because his
wife and children went away for a holiday in the country, while he was left in
town, working. The more recent fotballenke (wife left alone because her
husband is a football player) seems to have no corresponding fotballenkemann,
despite the fact that there are women football players. Finally, the term hanrei
‘cuckold’ is a term which has no feminine counterpart, obviously for the reason
that a woman’s faithfulness was important, and her lack of it merited a shame-
ful appellation for the husband. A man’s lack of faithfulness, on the other hand,
needed no special terminology for the betrayed woman.
Not only may women and men be referred to differently, but there are also
lingering and distinct differences with respect to how we talk about women and
men. Thus both women and men may be strong (sterke menn ‘strong men’,
sterke kvinner ‘strong women’), yet strong women often risk being called
dominerende ‘dominant’ or even maktgale ‘powermad’, while the men are
simply strong, or perhaps har autoritet ‘have authority’. At the beginning of her
career, Gro Harlem Brundtland (the previous Norwegian Prime Minister and
now Head of the World Health Organization) was regularly called kjeftesmella
fra Bygdøy ‘the shrewish chatterbox from Bygdøy’. The act of talking as such
seems to be gender-differentiated as well. Women are said to skvaldre, skravle,
skvatre (all meaning ‘chatter’), or sladre ‘gossip’. Corresponding nouns abound,
for instance: skravlekjerring ‘chatter woman’ (chatterbox), or sladrekjerring
‘gossipy woman’. Men’s talking is normally described as more sober, they are
said to snakke ‘talk’ or diskutere ‘discuss’.
In conclusion it can be said that although Norwegian has undergone quite a
few ideological feminist-inspired changes, asymmetries abound. However, there
is every reason to expect that some of these will change over time. On the other
hand, the fact that the language still may reflect a lack of equality even in a
country where gender equality in fact has come a long way, is rather depressing.
7. Language change and feminist language planning
The recent history of the Norwegian language is a story about language plan-
ning and language reform, and indeed written Norwegian, both bokmål and
nynorsk, has become very different today compared to the beginning of the 20th
century. Initially, there was no gender consciousness in these reforms; they
might, nevertheless, be regarded as sociolinguistic reforms, as they aimed at
eliminating social or class differences in language by giving former vernacular
forms the status of standard language.
All official language and gender interest has taken place over the last 20–30
years, as shown by the annual reports from the Norwegian Language Council and
the former Norwegian Language Commission.9 The first time any official language
planning agency or committee explicitly and by its own initiative put the issue
of language and gender on the agenda was in 1980 at the annual meeting of the
Norwegian Language Council (cf. Bull 1992). Prior to 1980, however, similar
questions had been raised by the Ministry of Church and Education.
The problems the ministry was interested in and wanted the Language
Council’s advice on were twofold: 1) how to achieve gender neutrality in
occupational and professional nouns and titles, such as formann ‘chairman’ and
fylkesmann ‘highest official of county’ and 2) whether new legislation on gender
Norwegian 245
equality should include the use of generic han ‘he’. For a long time, however,
the Language Council was not able to give the ministry any other advice than
just reiterating that people’s opinions differed.
In the 1980s, the question of whether to accept generic use of ho/hun ‘she’
in schoolbooks was raised on several occasions. The annual report of 1981
includes general guidelines which mention possible substitutions for the generic
masculine, but also advise the public and especially the authors of schoolbooks
to avoid using this pronoun when the referent is a woman, or includes women,
or when the gender is not known or relevant.
Gender-indefinite substitutes that were recommended were the pronoun
dei/de ‘they’, i.e. the plural form, the indefinite pronoun ein/en ‘one’ (as
opposed to indefinite man ‘one’, which is derived from mann), the noun folk
‘people’, or pluralisation of the noun phrase.
The public was advised not to use han/ho (or han/hun) or han eller ho/hun
‘he or she’; these expressions were said to be awkward, clumsy, and inelegant.
While the Council lamented the fact that Norwegian has no gender-indefinite
pronoun for the third person singular, they stated categorically that if a singular
pronoun is unavoidable, han was the only correct alternative in gender-indefi-
nite contexts, thus arguing that the masculine may comprise both masculine
and feminine. Ho/hun ‘she’ was acceptable only in cases where the context
showed that the referent was a woman (but then also with masculine nouns). If
the Norwegian equivalent of a sentence like The plumber must be ill today, she
has not turned up is found in a pre-1989 Norwegian schoolbook, we can be
quite sure that the plumber in question is a woman, and this means that generic
ho/hun ‘she’ actually was forbidden in Norwegian schoolbooks.
Thus the Language Council admitted that generic ‘he’ might cause prob-
lems for many Norwegians, but on the other hand they were unable to accept
that for some the opposite solution, generic ‘she’, was a way of making the very
problem visible.
In the years after 1980 the question of generic han was to haunt the Norwe-
gian Language Council. There are many reasons for this. Of course, the time
was ripe; the same questions of language and gender had been discussed
internationally since the mid-seventies. More specifically, however, in Norway,
the language as well as the content of all textbooks for use in school are subject
to monitoring by control boards (a practice that is becoming increasingly
controversial). The linguistic control is mostly a control of spelling, morpholo-
gy, syntax, as well as style, whereas the content of textbooks is monitored with
respect to rules advocating equality of women and men. Quite contrary to the
advice of the Language Council the members of the Equality Council urged
writers of schoolbooks to use what they called non-traditional pronouns as a
means to fight androcentrism and sexism. This precisely meant advocating the
use of generic ‘she’, which was forbidden by the Language Council. Under-
standably, the disagreement between the two Councils turned out to be quite a
delicate matter. In 1989, however, the Language Council again discussed this,
and by a majority decision, generic ‘she’ was finally accepted for use in child-
ren’s schoolbooks.
In many ways generic (or pseudogeneric) use of ho/hun ‘she’ is a marginal
problem, at least if we consider it quantitatively. Rather few make use of this
option, probably fewer within the context of schoolbooks than in newspapers
and periodicals. Still, it was a big issue during the years it was on the agenda of
the Language Council, and by feminists interested in the problem the result was
considered a great victory.
In 1994 the Language Council decided, if somewhat unenthusiastically, to
publish guidelines on non-sexist language. The result was the publication in
1997 of a small brochure with the title Gender, language and equality (Kjønn
1997) – seventeen years after the question of language and gender had been
raised for the first time. The brochure argues for the necessity of such guide-
lines, and then discusses fundamental rules for gender-balanced (inclusive)
language. Thus they advocate the use of ungendered words for names of
professions, and conversely non-use (or little use) of premodifying kvinnelig/
mannlig ‘female/male’ (as in kvinnelig tannlege ‘woman dentist’), except where
strictly necessary. The guidelines recommend that women and men should be
treated in the same way, and both should be made visible linguistically. The
examples given under this point include the use of han eller ho (the previously
despised ‘he or she’), and consistent use of pairs such as talsmann–talskvinne
‘spokesman–spokeswoman’. The brochure also warns against asymmetrical
pairs such as ‘men and girls’, and even advocates variation in the sequencing in
commonly used pairs, i.e. menn og kvinner ‘men and women’ and kvinner og
menn ‘women and men’, jenter og gutar ‘girls and boys’ as well as gutar og jenter
‘boys and girls’. All gender discrimination and stereotyping should be avoided.
The brochure points out that it is very common to use different expressions to
describe women and men: ‘a man of principle’, and ‘a stubborn woman’.
Finally, pronouns, titles and honorifics are also mentioned, and relevant and
adequate recommendations are given.
These guidelines are distributed to Norwegian schools and to public
institutions. There are, however, no reports on the use of the guidelines and
Norwegian 247
therefore no information about how successful they have been. In many ways,
it is fair to say that the guidelines appeared so late that they barely were able to
keep up with the factual language development. Thus, they may be looked upon
as descriptions of language use as much as prescriptive guidelines.
8. Conclusion
On the whole, then, official Norwegian guidelines follow the strategy of

gender-neutralisation rather than gender-specification, in line with what most
Norwegian feminists choose, and in line with what Hellinger (1984) calls the
“generic strategy”. Hellinger suggests that this strategy is typical of English as
well as Norwegian, while German-speaking countries favour a “visibility
strategy”. In other words, it has been the policy of Norwegian feminists as well
as political authorities to attempt to eradicate gender marking from the
language, except in those cases where specific persons are talked about.
Visibility has been less important, although some feminists have in fact
supported the “visibility strategy”. Thus, the use of pronominal splitting (han
eller ho) as well as the use of some explicitly female-specific terms, esp.
compounds with -kvinne as the second element, might perhaps be interpreted
as indications of a “mixed” strategy.
Even though it is quite obvious that the changes described above are
connected to ideology, the social aspects represent a very complex situation
with partly conflicting goals and values. Norwegian women have lost some
visibility in this process, but few people think of this as symbolic annihilation
(cf. Tuchman 1978). Strangely enough, ideologically based changes have
preceded language planning, or at least that of the Language Council.
Notes
1. Standard reference works of Norwegian are: Haugen (1965, 1976), Faarlund & Lie &
Vannebo (1997), Strandskogen & Strandskogen (1995), Vikør (1995).
2. Specification of lineage in terms for various relatives, such as moster ‘mother’s sister’ and
faster ‘father’s sister’ were more common in earlier stages of Norwegian.
3. Gender marking may also take the form of adding the adjective kvinnelig/mannlig
‘female/male’, as in kvinnelig dommer (m) ‘female judge’ or mannlig advokat (m) ‘male
lawyer’ – needless to say, in most professions the latter form would be unusual.
4. Actually the form norske (which is parallel to svenske ‘Swede’ and danske ‘Dane’) has been
suggested several times as a replacement. The Language Council, however, remains staunch
and calls this form “childish”.
5. Before the custom of proper inherited surnames gradually became normal usage in the
19th century – first in the upper strata of society – children might be called by their father’s
first name plus -sønn (later levelled to -sen) ‘son’ or datter ‘daughter’. Thus Per Pettersen
would be the son of Petter, and Randi Pettersdatter his daughter. The Name Act of 1923
regularised the use of inherited surnames.
6. As an example of variation we recently found a brief article about legefamilien Hans
Nielsen Hauge ‘the doctor family H.N.H.’ in a local paper. It seemed very odd, indeed, and
it would be totally impossible to create the female equivalent *legefamilien Nina Hauge.
However, a few days later the same paper carried an article about familien Tomter/Stigen (i.e.
a family identified with reference to the surnames of both spouses), and this certainly is the
normal practice in present-day Norwegian. Nevertheless, it should be noted that to its (?her)
shame, Norway still considers the man the head of the family for tax purposes, while the
home in a sense is defined as the woman’s domain. Thus married couples are taxed
individually, but the wives are always filed under the husband’s birth date and name. Married
couples must also by law be registered in the same municipality, even if they live and work
in different cities; however, for tax purposes, the home community is where the woman lives.
7. Prestefruen was not just a title, but a concept in itself. Needless to say, the traditional
prestefrue has all but disappeared – her job was the unsalaried, but full-time, task of taking
care of the prest as well as the parish.
8. This is still the case as far as the Norwegian monarchy is concerned, while in Sweden, the
heir to the throne is a woman, even though she has a younger brother.
9. The Norwegian Language Council, which among other things is in charge of the language
of Norwegian schoolbooks, was founded in 1972, replacing the Language Commission of
1952–1971. The Council has 38 members, and is divided into two sections, one each for
nynorsk and bokmål. The members are appointed by different institutions and organisations:
the universities, the teachers’ organisations, authors’ organisations, the national broadcasting
system, the theatres, private organisations of language politics, etc.
References
Beekes, Robert S. P. 1995. Comparative Indo-European linguistics. Amsterdam: Benjamins.

Beito, Olav. 1986. Nynorsk grammatikk [Nynorsk grammar]. Oslo: Samlaget.
Blakar, Rolf Mikkel. 1973. Språk er makt [Language is power]. Oslo: Pax.
Blakar, Rolf Mikkel. 1975. “How sex roles are represented and conserved in the Norwegian
language.” Acta Sociologica 18: 162–173.
Bull, Tove. 1985. “Apropos enke/enkemann” [Apropos ‘widow/widower’]. Språknytt 13:
15–16.
Bull, Tove. 1990. “Til mann og kvinne skapte han dei” [Man and woman He created them].
Altaboka. Alta: Alta Historielag, 89–95.
</TARGET "bul">
Norwegian 249
Bull, Tove. 1992. “Male power and language planning: The role of women in Norwegian
language policy?” International Journal of the Sociology of Language 94: 155–171.
Ehrlich, Susan & Ruth King. 1994. “Feminist meanings and the (de)politicization of the
lexicon.” Language in Society 23: 59–76.
Eide, Elisabeth. 1993. “Latterlig eller usynlig?” [Ridiculous or invisible?]. In Dagens dame – fri og
likestilt? [The lady of the day – free and equal?], ed. Brit Fougner. Oslo: Aschehoug, 59–80.
Faarlund, Jan Terje & Svein Lie & Kjell Vannebo. 1997. Norsk referansegrammatikk [Norwe-
gian reference grammar]. Oslo: Universitetsforlaget.
Haugen, Einar. 1965 Norwegian English dictionary/Norsk engelsk ordbok. Madison: University
of Wisconsin Press & Oslo: Universitetsforlaget.
Haugen, Einar. 1976. The Scandinavian languages. An introduction to their history. London:
Faber & Faber.
Heggstad, Kolbjørn 1982. Norsk frekvensordbok [Norwegian frequency dictionary]. Ber-
gen: Universitetsforlaget.
Hellinger, Marlis. 1984. “Effecting social change through group action: Feminine occupa-
tional titles in transition.” In Language and power, eds. Cheris Kramarae & Muriel
Schulz & William O’Barr. London: Sage, 136–153.
Jahr, Ernst Håkon, ed. 1995. Sociolinguistics in Norway. Berlin: de Gruyter.
Kjønn 1997. Kjønn, språk og likestilling [Gender, language, and equality]. Oslo: Norsk språkråd.
Landfald, Aagot & Kjell Paulsen, eds. 1996. Norsk ordbok [Norwegian dictionary]. Oslo:
Cappelen.
L’Hoir, Francesca Santoro. 1992. The rhetoric of gender terms. Leiden: Brill.
Romaine, Suzanne. 1999. Communicating gender. New York: Erlbaum.
Ryen, Else, ed. 1976. Språk og kjønn [Language and gender]. Oslo: Novus.
Schulz, Muriel R. 1975. “The semantic derogation of woman.” In Language and sex:
Difference and dominance, eds. Barrie Thorne & Nancy Henley. Rowley, MA: Newbury
House, 64–75.
Strandskogen, Åse-Berit & Rolf Strandskogen. 1995. Norwegian. An essential grammar.
London: Routledge.
Swan, Toril. 1991. “Hvor synlige var kvinner i norsk presse omkring 1911?” [How visible
were women in the Norwegian press around 1911?]. In Den lange veien til parnasset [The
long way to Parnassus], ed. Åse Hiorth Lervik. Tromsø: University of Tromsø, 82–91.
Swan, Toril. 1992. “All about Eve: Women in Norwegian newspapers in the 20th century.”
Working Papers on Language, Gender and Sexism: 37–54.
Swan, Toril. 1995. “Gender and language change: The case of Norwegian.” Nordlyd 23: 298–315.
Tuchman, Gaye. 1978. “The symbolic annihilation of women by the mass media.” In Hearth
and home. Images of women in the mass media, ed. Gaye, Tuchman. New York: Oxford
University Press, 3–38.
Vikør, Lars. 1995. The Nordic languages. Oslo: Novus.
<TARGET "nis" DOCINFO
AUTHOR "Uwe Kjær Nissen"
TITLE "Gender in Spanish"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
spanish
Gender in Spanish
Tradition and innovation
Uwe Kjær Nissen

University of Southern Denmark, Odense, Denmark
1. Introduction
2. Grammatical gender in Spanish
2.1 Masculine and feminine gender
2.2 Generic words and conflicts of agreement
2.3 The psycholinguistic reality of generics
3. The asymmetrical representation of women and men in Spanish
3.1 Hombre vs. mujer
3.2 Occupational titles
3.3 Sexual terminology
3.4 Sexism in selected text types
3.4.1 School books
3.4.2 Dictionaries
5. Guidelines for a non-sexist use of Spanish
6. Conclusion
Notes
References
1. Introduction
A member of the Romance family of languages, Spanish (Español) is the official

language of Spain, almost all of Central and South America (except Brazil,
Suriname and Guyana), and Equatorial Guinea; it is spoken by over 300 million
people. A large number of first-language and bilingual speakers of Spanish live
in the United States, Puerto Rico, Western Sahara and the Philippines. Al-
though a number of local varieties exist due to regional influences, in general
252 Uwe Kjær Nissen
two main “standard” norms are recognised: Peninsular (also called “Castilian”,
originally the dialect spoken in the kingdom of Castile) and Latin American.
Despite some morphosyntactic differences, mutual intelligibility between both
is high. One reason for this uniformity is the work of La Real Academia de la
Lengua Española (‘The Royal Academy of the Spanish Language’), located in
Madrid and founded in 1713. Up until today, this authoritative and, more often
than not, conservative Academy has the last word on linguistic issues that deal
with Spanish.
Typologically, Spanish is a VO language; sentences need no overt subject,
as this is traceable by means of verb morphology. When there is an overt
subject, SV(O) is the rule, although this order is never obligatory. In general,
modifiers (i.e. adjectival phrases and relative clauses) follow their syntactic
heads, but some common adjectives precede them. Spanish is considered a
synthetic rather than analytic language and its derivational processes predomi-
nantly use suffixation, although, especially in recent times, prefixation has been
slightly increasing (Bright 1992: 61). Apart from the verb system, inflection is
found in pronouns, nouns, adjectives and determiners, where it marks number
(i.e. plural -s) and gender (i.e. -o / -a), although – with respect to person
reference – it has been debated whether the latter is, in fact, a derivational rather
than inflectional process.
In the course of history, Spanish has been influenced by a variety of other
languages, such as Arabic, Basque, Catalan, French, German, Italian, and
English as well as – after the colonisation of the “New World” – indigenous
languages of Latin America (e.g., Nahuatl, Quechua and Arawak).1
Linguistic studies that treat aspects of “gender and language” are, in the
majority of cases, either fostered in Spain or concentrate upon the Castilian
variant. Research into this area from Latin America is scarce (cf. nevertheless,
note 18). One possible explanation for this difference is the rapidity of social
changes in post-Franco Spain influenced by European politics after Spain
joined the European Common Market in 1983. Moreover, at the beginning of
the 1980s Spain became more aware of women’s issues and experienced a
growing feminist movement, which was strongly backed up by initiatives of the
Socialist Party that won the general elections in 1982. For example, the Instituto
de la Mujer ‘The Woman’s Institute’ was founded in 1983 as part of the Minis-
try of Science and Education (now the institute belongs to the Ministry of Social
Affairs). At the same time, an increasing number of women joined the labour
market and entered educational institutions. Naturally, these changes in society
have had a considerable impact on the Spanish language; difficulties in finding
Spanish 253
appropriate job titles for women are a paramount example of this. This article
will concentrate on four main areas: grammatical gender (Section 2), the
asymmetrical representation of women and men in the lexicon and in language
reference material (Section 3), the communicative behaviour of women and
men2 (Section 4), and non-sexist guidelines and language change (Section 5).
For further reading on language and gender in Spanish, consult Calero Fernández
(1999), García Mouton (1999), and Fernández de la Torre Madueño & Medina
Guerra & Taillefer de Haya (2000).
2. Grammatical gender in Spanish
2.1 Masculine and feminine gender

Traditionally, the concept of gender in Spanish was defined on the basis of both
semantic and morphological criteria of the noun class. Thus, the first grammar
book on Spanish, written by Antonio Nebrija more than five centuries ago
(Nebrija 1492), distinguished six different genders. Unfortunately, reminiscenc-
es of this classification have survived until today and the Academy, for example,
still classifies nouns in its dictionary into three groups: m. for masculine, f. for
feminine and com. for común (= those nouns that are both m and f), which
apparently lead even experts on this matter to conclude that “Spanish has three
genders for persons, the masculine, the feminine and the common” (García
Meseguer 1977: 137).3 A more convincing approach seems to be that of Corbett
(1991). According to him, a noun (in his terminology the “controller”) induces
the same gender in other linguistic elements (the “targets”). As all targets in
Spanish, e.g., determiners, adjectives, and predicates, vary morphologically
according to the gender of the nouns, it is quite straightforward to conclude
that the Spanish noun class has no more than two genders, namely masculine
and feminine. However, at times, it appears convenient to categorise the nouns
according to other criteria. Thus, one specific group of nouns in Spanish is
invariant as to morphological changes due to gender, and, in isolated cases,
where no modifier is present, one cannot tell whether the noun is feminine or
masculine. For example, all words that end in -ista belong to this group and,
therefore, e.g., marxista may refer either to a woman or a man. On the other
hand, as both genders are intrinsic properties of these nouns, they are, when a
gender-marking target is added, instantiated as either masculine or feminine –
in the majority of the cases according to the referential gender. For these
reasons, the term común to indicate gender is not well-chosen and should be
avoided. A more appropriate term is double gender (cf. Corbett 1991: 181).
Therefore, in this article, nouns like periodista ‘journalist’, modelo ‘model’,
policía ‘police officer’, etc. are referred to as nouns of double gender.4
As gender is defined on the basis of agreement, special focus is placed on
agreement-markers. For Spanish these are, in most cases, -o (m) and -a (f).
Apart from some minor deviations (where the opposition, if it exists, is Ø vs. -a
or -e vs. -a), most adjectives, determiners and pronouns follow this vowel
distinction. With respect to the gender of the controllers, one also finds
remarkable consistency, which underlines the morphosyntactic functioning of
the Spanish gender system: 99.89% of the nouns ending in -o are masculine,
96.6% of the nouns ending in -a are feminine, and 89.35% of the nouns ending
in -e are masculine (Teschner & Russel 1984: 116–117). Although this categor-
isation appears to be formally motivated and semantically arbitrary for the
noun class as a whole (e.g., libro vs. libra ‘book’ vs. ‘pound’), this is not the case
for nouns that refer to animate entities.
If generic references and so-called epicene nouns (see Section 2.2) are not
taken into consideration, feminine and masculine nouns referring to human
beings in the majority of cases reflect the distinction between females and males
(a few counter-examples exist that constantly reappear in the literature about
this topic).5 In these cases the morphological distinction follows the same line:
abogado vs. abogada ‘lawyer’, profesor vs. profesora ‘teacher’ (i.e. non -a vs. -a).
That this pattern is strongly associated with ‘female’ and ‘male’ can also be
deduced from examples that refer to animals where referential gender is not
usually expressed. For example, an invariable noun6 such as el tigre (m) ‘the
tiger’ would prompt a word like la tigra (f) or la tigresa (f), if speakers of
Spanish were to make a distinction between female and male tigers. Metaphori-
cal use also reflects the relation between ‘female’ and feminine, ‘male’ and
masculine: Inanimate words like e.g., la muerte (f) ‘death’ will usually be
portrayed by a female figure (by contrast, speakers of German associate der Tod
(m) with a male figure). This association of morphology with the categories
‘female’ and ‘male’ is very deep-rooted, as can be demonstrated by the following
example from the inanimate world. In 1977, the feminist journal Vindicación
Feminista ridiculed another journal, Telva, which at that time was against the
legalisation of abortion. The feminist journal urged Telva to change its name
and slogan from Telva, revista para mujeres ‘Telva, journal for women’ to Telvo,
revisto para mujeros. By means of this “masculinisation” the journalists wanted
to emphasise that Telva was no longer a journal for women, but had taken the
Spanish 255
opposite, anti-female/male position on this much debated issue (Vindicación

Feminista 18, 1977: 49).
There can hardly be any doubt that the endings -o vs. -a (or, rather, non -a
vs. -a) linguistically serve as markers of maleness and femaleness when humans
are involved and that these endings are also perceived as such markers by
speakers of Spanish.7
An inconsistency with regard to the graphic representation of these endings
is found in many dictionaries, in which feminine words are presented as if they
were simply derived from the masculine forms, e.g.: niño/-a ‘boy/girl’.8 Obvi-
ously, this cannot be the case. A more appropriate procedure would be to write:
niñ/-o, -a or niñ/-a, -o.9 That this matter is not trivial can, for example, be seen
in etymological dictionaries: These usually list only the masculine forms.
Feminine forms and, hence, women, have no origin (a welcome exception is
Müller 1987–2000).
Although today the suffix -a is the predominant productive derivational
suffix, a few others still exist: -isa, -esa, -iz, -ina, -sa, all of which are feminine
suffixes. Interestingly, Spanish has no masculine derivational suffix (if we
discount instances of the type niñ-o, where a masculine marker is attached to
the stem of a word). Whereas English, e.g., has -er, as in widow Æ widower, and
likewise Danish has -mand as in enke ‘widow’ Æ enkemand ‘widower’, Spanish
has nothing of this kind. In those cases where a masculine noun, for historical
reasons, has “derived” from a feminine noun, as in enfermero from enfermera
‘nurse’, synchronically this can hardly be traced.10
Gender-marking by means of compounding does not exist; however, in
those cases in which a derived word would be considered unusual, lexical
modifications are used. Primarily the noun mujer ‘woman’ is used as a pre-
modifier: e.g., mujer (f) obispo (m) ‘female bishop’ or mujer (f) torero (m)
‘female bullfighter’ instead of the morphologically possible derivations obispa
(f) and torera (f).
A fundamental area in which gender distinctions, for various cultural
reasons, are prominent, is kinship terms. While many European languages
distinguish the gender of family members lexically, Spanish (like other
languages derived from Latin) is characterised by alternating vowels to
distinguish referential gender (cf. Table 1). Cf. also sobrina/sobrino
‘niece/nephew’ (exceptions are madre/padre ‘father/ mother’ and nuera/yerno
‘daughter-in-law/son-in-law’).
One feature Spanish shares with many (probably most) other languages is
the fact that the masculine form functions as the so-called “unmarked” form,
Table 1.Kinship terms in some European languages

English daughter/son sister/brother aunt/uncle
German Tochter/Sohn Schwester/Bruder Tante/Onkel
Danish datter/søn søster/bror tante/onkel
Spanish hija/hijo hermana/hermano tío/tía
i.e. when reference is made to a mixed group, the masculine has to be used.
Even though many reference books of Spanish – and even publications that
specifically address gender issues – treat this phenomenon as if it only applied
to the plural, it is important to stress that the masculine also functions as the
unmarked form in the singular. Thus, in a sentence such as (1) the singular
form el obrero is supposed to cover female workers as well:
(1) El obrero (m) del siglo pasado trabajaba mucho.
‘The worker in the last century worked a lot.’
In principle, there is no difference (apart from stylistic variations) between this

sentence and its plural version: Los obreros (m) del siglo pasado trabajaban
mucho. In both cases the masculine form is ambiguous. No empirical study
exists that specifically investigates which of the two is more inclusive.
2.2 Generic words and conflicts of agreement

Traditionally the problem of generic reference, i.e. referring to women and
men at the same time in such a way that both the speaker and the hearer
interpret the message as referring generically to human beings, has been
related to the concept of markedness, the masculine form being the unmarked
term. However, usually no explanation is given why it is the masculine and not
the feminine that became the unmarked term. Lyons raised doubt about the
concept of markedness by introducing a model of “dynamic marking” (Lyons
1977: 305–311). He observed that some words in the same contexts are more
marked than others. The following Spanish sentences will exemplify his
observation. In (2a) and (2b) the word trabajadores (m) ‘workers’ is used as the
unmarked term:
(2) a. Juan (m) y Pedro (m) son trabajadores (m) extraordinarios (m).
‘Juan and Pedro are extraordinary workers.’
b. Juana (f) y Pedro (m) son trabajadores (m) extraordinarios (m).
‘Juana and Pedro are extraordinary workers.’
Spanish 257
However, by exchanging trabajadores with another supposedly unmarked (and

hence generic) term, hombre ‘man’, (3b) becomes inacceptable:
(3) a. Juan (m) y Pedro (m) son hombres (m) extraordinarios (m).
b. *Juana (f) y Pedro (m) son hombres (m) extraordinarios (m).
*‘Juana and Pedro are extraordinary men.’
This suggests that in Spanish, too, “marked” and “unmarked” are not discrete
categories, but should, instead, be viewed as instances on a scale (cf. Lyons
1977, Hellinger 1990).
In the domain of genericity, some terminological clarifications are required.
Traditionally, both single gender nouns like e.g., personas (f) ‘persons’ and
masculine nouns used generically, like e.g., padres (m) ‘parents, fathers’ are
called “epicene” nouns (Real Academia Española 1977: 175–176), supposedly
because both forms may refer to women and men. It is, nevertheless, important
to distinguish these two forms, as it is only the first group of nouns that can be
claimed to be neutral. A neutral expression11 should be capable of referring to
women and men, to men alone and to women alone (cf. Pusch 1980: 181). With
epicene nouns this is possible, cf. (4):
(4) a. Juan (m) y Juana (f) son personas (f) extraordinarias (f).
‘Juan and Juana are extraordinary persons.’
b. Juan (m) es una (f) persona (f) extraordinaria (f).
c. Juana (f) es una (f) persona (f) extraordinaria (f).
‘Juan/-a is an extraordinary person.’
In comparison, this is not the case with the masculine form:

(5) a. Juan (m) y Juana (f) son campesinos (m) extraordinarios (m).
‘Juan and Juana are extraordinary farmers.’
b. Juan (m) es un (m) campesino (m) extraordinario (m).
c. *Juana (f) es un (m) campesino (m) extraordinario (m).
‘Juan/-a is an extraordinary farmer.’
In general, masculine nouns for which a corresponding noun of the feminine

gender exists, cannot refer to women alone, simply because a corresponding
noun is available. Therefore, the term “epicene” should only be used when no
counterpart of the opposite gender exists. This is true for both subgroups of
epicene nouns: a) collectives such as gente (f) ‘people’, población (f) ‘population’,
profesorado (m) ‘teaching staff’ and b) non-collectives like persona (f) ‘person’,
individuo (m) ‘individual’, bebé (m) ‘baby’. In all these cases, the term can be
applied to a feminine and a masculine noun alike. An advantage of this proce-

dure is that the concept of markedness becomes irrelevant for epicene nouns.
To distinguish the masculine form from the epicene is not a trivial matter,
as can be seen from the reaction of one of the Academy members towards the
first guidelines for non-sexist Spanish, in which it was proposed that the use of
the masculine form for generic reference be avoided. He asks whether, if one
has to avoid “epicene words like romanos (m) ‘the Romans’, hispanos (m) ‘the
Hispanics’ and profesores (m) ‘the teachers’, should one not also avoid feminine
words like la víctima ‘the victim’ and la persona ‘the person’ as well?” (Lázaro
Carreter 1988: 10). Naturally, the answer is no. In the latter cases no counterpart
of the other gender exists: *el víctimo, *el persono12 and, subsequently, no item
is unmarked (nor marked for that matter). Therefore, there is no reason to
avoid them on the grounds of presumed sexism. Masculine words, on the
contrary, function differently: When referring to human beings, the masculine
word will automatically be unmarked and the feminine term marked.13
As the generic masculine, in contrast to the epicenes, may refer to men and
women as well as to men alone, the term in many cases appears to be a mislead-
ing denomination. Therefore, I prefer, simply, to use the term “ambiguous
masculine”.
One consequence of the masculine form being the unmarked term has to
do with anaphora and agreement. For example, one single man in an audience
will be sufficient for the speaker to address the listeners by masculine terms,
e.g., vosotros (m) ‘you’ or los presentes (m) ‘the ones present’. To use feminine
forms in such cases would be tantamount to breaking the rules of grammar.
The same is true with respect to agreement. If a masculine word and a feminine
word are co-ordinated, the modifying elements of the head have to be mascu-
line. In some instances, when the nouns refer to animals or to inanimate
entities, this general rule is suspended. For example, when an adjective is placed
before two co-ordinated nouns of which the first is feminine and the second
masculine, the adjective may agree with the feminine noun (= agreement
between those elements that are closest). But, strikingly, when the nouns refer
to human beings, this principle is not followed and the general rule is applied
again (cf. also Marcato & Thüne, this vol.) on the corresponding phenomenon
in Italian). Apparently, men cannot “be agreed with” by feminine modifica-
tions.14 The most plausible explanation for this phenomenon appears to be
Spender’s “man-made-language” approach (1980).
Spanish 259
2.3 The psycholinguistic reality of generics

Despite the fact that the masculine form overwhelmingly seems to dominate the
structure of Spanish and, consequently, is claimed to make women linguistically
less visible (see Suardiaz 1973, as well as Section 5), there are few empirical
investigations which specifically address the question of how speakers interpret
these forms.
One of the first investigations that dealt with Spanish generics set out to
analyse the generic potential of the word hombre ‘man’ (Perissinotto 1983). The
tests showed that most women and men in nearly all contexts interpreted
hombre as specifically male. Another study by Perissinotto (1985) also proved
hombre to be a poor generic. On a hierarchical scale informants placed hombre
closest to the male-specific interpretation, whereas e.g., gente (f) ‘people’ and
persona (f) ‘person’ were interpreted generically. Other epicene words, such as
individuo (m) ‘individual’ and ser humano (m) ‘human being’ were placed in
between these two poles.
The point of departure of the most recent study of this kind (Nissen 1997a)
was the various guidelines for non-sexist language (cf. Section 5) which
recommended that the ambiguous masculine form be replaced by other forms
in order to assure a generic interpretation. Based on a blind test, three different
types of almost identical questionnaires were distributed among some 300
native speakers in Spain, the majority of whom were students. One type
contained forms of the ambiguous masculine, e.g., los niños (m) ‘the children’,
the second one epicene forms, e.g., la población infantil (f) ‘the population of
children’, and the third contained splitted forms, e.g., los niños (m) y niñas (f)
‘the boys and girls’. The participants were asked to insert proper names in the
text following these forms. As there were always two gaps, responses could be
entirely male (two male names), entirely female (two female names) or mixed
(one female and one male name). As had been expected, the results show that
the male bias is strongest in relation to the masculine form: In approximately
one third of the cases, it is associated with male (i.e. only male names were
used). In the case of splitted forms, this association is only made in one fifth of
the occurrences. Furthermore, the splitted forms seem to provoke a stronger
“generic thinking” in comparison to the masculine form (i.e. more mixed name
pairs) and also a stronger association with female (i.e. higher percentage of
female-only responses) so that the visibility of women increases. Surprisingly,
and contrary to what had been expected, the epicene forms are not interpreted
generically to a higher degree than the masculine forms, since the percentage of
“mixed” responses is about the same in both cases. When focus is placed on the
gender-specific interpretations alone, another surprising result emerges: Both
the masculine and the epicene forms are slightly male-biased, but the splitted
forms are biased towards female (i.e. more female-only than male-only respons-
es). Hence, if the aim is to make women more visible, this splitted construction
is definitely the best candidate.
If the gender of the respondents is taken into account, the survey shows
that, although for both women and men the visibility of women has increased
with the use of the splitted forms (instead of the masculine), this increase is
much higher for men. In a way, it seems as if men have to be “reminded” of the
existence of women (by means of the splitted forms). On the other hand, all
respondents who did not connect the splitted forms with genericity (almost
10%) did associate them with their own group (despite the fact that both
females and males were explicitly mentioned).
To summarise, the investigation demonstrates that alternative forms to the
masculine increase the visibility of women, but although the masculine forms
indeed had a slight bias towards male, the claim that these forms automatically
lead people to think ‘male’ was not substantially confirmed. It is nevertheless
true that the masculine form is far from being a true generic, and certainly the
shape of a word – and thereby its grammatical gender – does play a considerable
role in its interpretation. Since this survey focused only upon certain nouns in
isolated sentences, no information is available as to how these forms would be
interpreted in texts, nor whether other gender marking devices interfere in this
interpretation (e.g., in connection with pronoun substitution).15
3. The asymmetrical representation of women and men in Spanish
Lexical asymmetries in the Spanish lexicon with respect to human beings can be
classified in the following way:
a. One type includes different designations for women and men, especially
with respect to job titles: azafata ‘stewardess’ versus auxiliar de vuelo ‘flight
attendant’. Moreover, apparently equivalent words such as e.g., señor ‘Mr,
gentleman’ versus señora ‘Mrs, lady’ are not always symmetrical on a
denotational level. For el señor del tango ‘master of the tango’ the equivalent
is la reina del tango ‘queen of the tango’; *la señora del tango does not exist.
b. A second type shows semantic differences of the same lexical item, depend-
ing on whether the item refers to a man or a woman: hombre público ‘man
Spanish 261
in the public eye’ versus mujer pública (traditionally) ‘whore’.

c. Thirdly, there are lexical gaps: ama de casa ‘housewife’ (*amo de casa does
not exist); soldado ‘soldier’ (*soldada ‘female soldier’ is rarely found);
señorita ‘Miss’ (the word señorito does exist, but does not mean ‘unmarried
young man’, but ‘young gentleman, master, rich kid’).
3.1 Hombre vs. mujer

With regard to the asymmetries as in group (b), the word hombre ‘man’ has
been particularly subject to semantic analysis (García Meseguer 1984, Bierbach
1989). Two areas have received special attention: one, where hombre stands in
contrast to mujer ‘woman’ and two, where hombre is used generically (cf.
Section 2.3). The opposition hombre/mujer is highly asymmetrical in Spanish,
as mujer more often than not contains implicit derogatory connotations that
range from ‘sexual object’ to ‘prostitute’. It has been suggested that the word
mujer has been constantly “contaminated” by the meaning of ‘prostitute’
(Fuertes Olivera 1992: 122). During the Gulf war, a very influential newspaper
in Spain, El País, wrote:
(6) Los soldados (m) estadounidenses en el frente echan de menos el alcohol y
las mujeres (f).
‘The American soldiers at the front are longing for alcohol and women.’
(El País, 26 January 1991: 8; cited by Fuertes Olivera 1992: 122)
Considering that female soldiers actually were present during that war, the
example suggests a connection of mujeres to “sexual encounter”. (Note that the
masculine form los soldados cannot be interpreted generically). Other stereotyp-
ical connotations have also been noted. For example, an expression like actuar
como un hombre ‘to act like a man’ will focus on strength, self-assertiveness and
bravery, while actuar como una mujer means, when applied to a man, to be
weak, devious and cowardly (Suardiaz 1973: 56).
Another interesting point is the fact that mujer also means ‘wife’ (in more
formal settings one would use esposa). In comparison, hombre does not mean
‘husband’ (except in vulgar language); instead, there exists a specific word:
marido. To avoid confusion between the meanings ‘wife’ and ‘woman’,
definiteness may help in some cases: tiene mujer means ‘he has a wife’ (i.e. ‘he
is married’) whereas tiene una mujer means ‘he has a woman’ (who certainly is
not his wife).
3.2 Occupational titles

Occupational titles for women and men seem to constitute an especially
sensitive part of the vocabulary, as these titles are used to identify ourselves and
others. To have the “correct” title is, therefore, not a trivial matter. For example,
one or two decades ago, Spanish women did not agree upon whether a female
minister should be called ministra (f) or ministro (m). Arguments in favour of
the latter are, for example, the derogatory meaning of some feminine expres-
sions (e.g., verdulera means not only a ‘female vegetable seller’ but also ‘a
shameless woman’) and the possible reading ‘wife of ’ (e.g., médica means
‘female doctor’ but also ‘wife of the doctor’). After numerous disputes over this
matter, not only in the media but also in linguistics circles (cf. the discussions
between Olivares 1984, 1986, and Nissen 1986), opinion seems now to have
settled clearly in favour of the feminine form.
Today it is difficult to find examples in newspapers and magazines such as
la (f) ministro (m) ‘the female minister’ or similar constructions that were quite
normal only 20 years ago. All guidelines for non-sexist Spanish recommend the
use of the feminine form (cf. Section 5), and so do the editorial style manuals of
the leading newspapers (cf. Lebsanft 1997). The Royal Academy also advocates
the use of the feminine forms, and has, in fact, done so for many years. As early
as 1947 one member of the Academy stated the Academy’s view: The syntactic
clash between a feminine determiner and a masculine noun as e.g., in the title
la (f) catedrático (m) ‘the professor’ “is absolutely impossible to swallow”
(Casares 1947: 303). This position was backed up by various members of the
Academy years before women joined the labour market in massive numbers (cf.
Berro-García 1952/53 and Bustamente y Rivero 1960). However, despite the
Academy’s proclaimed intention to promote feminine job titles in the dictio-
nary, it continues to be a much debated puzzle as to why this task is still today
far from being realised (see also Section 3.4).16 Indeed, the Academy has
suffered much criticism for not being up-to-date with the occupational titles in
its dictionary and not being consistent in the explanations of the terms it gives.
Interestingly, due to the apparent reluctance of the Academy to make
changes, a member of the Spanish Congreso in 1986 proposed a motion to
abolish all sexist formulations in the Academy’s dictionary, but her motion was
not approved. However, since institutions, especially in the educational sector,
continued to be reluctant to introduce the use of feminine titles, the Ministry of
Education and Science decreed in March 1995 that 21 academic titles should
officially be feminised if the title holder was a woman (Errazu Colás 1995: 81).
Spanish 263
Thus, the long tradition of giving women academics their university diploma
with masculine titles, e.g., doctor instead of doctora, is in the process of disap-
pearing. Even those women who had already received a ‘male’ version of their
diploma are now entitled to receive a new, feminised one.
Although today, in the majority of cases, the feminine forms are used in
specific reference to females, this tendency sometimes appears to be overshad-
owed by a few examples that – for various morphological and phonological
reasons – are reluctant to feminise and, therefore, constantly reappear in the
discussions about feminisation. For example, currently both a noun of double
gender, el/la juez ‘the judge’, and a feminised version, la jueza, co-exist. Words
ending in -ante, as in estudiante ‘student’, comerciante ‘merchant’ do not
change into a feminine form as easily as others. However, especially in Latin
American Spanish, la estudianta is used more frequently. Only ten years ago,
the form la presidenta was used to refer to the chairwoman of associations,
clubs, etc., while la presidente referred to the female president. Today presi-
denta is used in both cases.
Another special case are occupational titles for men who take over jobs
traditionally held by women. Often an entirely new lexeme is introduced, and
sometimes it is even accompanied by an “upgrading” of the occupation
designated by the masculine term. In this respect, the word modisto constitutes
an interesting example, because modista originally belonged to the group of
nouns of double gender (as all nouns ending in -ista do). However, the word
modista ‘seamstress’, despite being an invariable noun, was transformed into
modisto, now meaning a fashion designer (cf. Nissen 1986: 735–736). Similarly,
auxiliar de vuelo, lit. ‘flight assistant’, may connote a higher status than the
morphological counterpart of azafata (f) ‘stewardess’, namely azafato (m).
3.3 Sexual terminology

In Section 3.1, the word mujer, meaning ‘prostitute’, was discussed. When
regarding women as sexual objects, numerous other terms which usually denote
females may also be “contaminated”: e.g., chica ‘girl’, as in chica de alquiler ‘girl
to hire’ or chica fácil ‘easy girl’. One has to bear in mind, however, that given the
change in gender roles, a few words exist today that show that men may also be
regarded as sexual objects, e.g., the word tío (lit. ‘uncle’) may be used by women
to denote a man as a sexual object, and the same counts for macho ‘male’ and
chico de alterne ‘boyfriend to be exchanged’ (Fuertes Olivera 1992: 126).
In accordance with the stereotypical picture of Hispanic culture as typically
machista, descriptions related to sexuality portray men as the active part and
women as passive. As Suardiaz (1973: 27) has pointed out, “[…] there are two
kinds of verbs which refer to intercourse: those which accept only male subjects
or both female and male subject”. No verb takes exclusively female subjects.
Thus, cogieron ‘they fucked’ (in Latin America, coger usually means ‘fuck’, in
Spain rather ‘seize, hold on to’) is seen from a joint perspective, while él la cogió
‘he fucked her’ and also the so-called ethical dative construction él se la cogió
(where the male is not only the agent of the action, but also the beneficiary) are
seen from a male perspective. Coger with a female subject does not exist. Even
neutral expressions, as e.g., hacer el amor ‘to make love’, which from a structural
point of view allow a female, male or a plural subject, still sound odd with a
female agent: Ella le hizo el amor (a él) ‘She made love to him’.
Another characteristic that Spanish shares with other languages is the lack
of a female counterpart to the word potente ‘potent’ referring to virility.
Although for men there exists a semantic opposition, impotente/potente, this is
not the case for women: frígida/Ø ‘frigid’. There are only few terms denoting
the sexual capabilities of women with either neutral or positive connotations.
Another aspect of this “double standard” is reflected by word pairs such as
hombre honrado / mujer honrada. The masculine expression simply means
‘honest’, whereas the feminine term may denote that the woman does not have
sexual relations except with her husband. Furthermore, the traditional concep-
tion of the untouchableness of the mother figure can be exploited negatively:
hijo de puta ‘son of a whore’ or la puta que te parió ‘the whore who gave you
birth’ are highly insulting expressions in Spanish. The very typical Mexican
expression chinga a tu madre ‘fuck your mother’ fits into this category (cf. also
the Turkish parallels described by Braun, vol. I).17
The metaphorical use of terms referring to male genitals in Spanish can be
interpreted as reflecting positive evaluations (Suardiaz 1973). It is well known,
for example, that tiene cojones ‘he has balls’ is widely used (in very informal
speech) to denote ‘brave’. Surprisingly, this expression can also be used nowa-
days for women: Thus, young electoral supporters backed up the female candi-
date for presidency in Colombia, Noemí Sanín, by slogans such as: ¡Noemí
tiene cojones! (Cambio 16, 29 June 1998: 64). For comparison, equivalent (?)
metaphorical expressions of approval such as *tiene ovarios ‘she has ovaries’ or
*tiene tetas ‘she has tits’ do not exist (and even if they did, they would not be
applied to men).
Spanish 265
3.4 Sexism in selected text types

3.4.1 School books
When discussions about gender and language emerged at the beginning of the
1980s in Spain, school textbooks were one of the first text types to be analysed
for alleged sexism.18 As many schoolbooks at that time had either been pub-
lished during the Franco period or had been written by authors that were
brought up during that time, sexist and stereotypical examples were frequently
encountered – not only in the texts, but also in the illustrations used. The fact
that many schools during this period were only for boys or for girls and were
now gradually being “merged” (the key term for this was coeducación), contrib-
uted to focussing on issues of sexism in the textbooks. However, even by 1992,
textbooks published between 1987 and 1989 on language and literature for use
in secondary school revealed surprising data about the representation of women
and men: In all the texts that were analysed, women appeared only in 2.9% of
the cases and men in 97.1%. In illustrations, women fared slightly better: 19.3%
as compared to 80.7%. One book about Spanish literature mentions only one
female author, but 79 male writers (Lledó & Otero 1994:374–375). Seen from this
perspective, it comes as no surprise that the first official guidelines for non-
sexist Spanish were published by the Ministry of Science and Education and
distributed to teachers and authors of schoolbooks (cf. Section 5).19
3.4.2 Dictionaries
Another object of study has been dictionaries, both mono- and bilingual. These,
together with grammar books, are of special interest, as they are often drawn
upon to settle disputes about language matters and serve as guides for “correct”
language use. Although lexicographers’ sexist presentation of language material
has often been confused with sexism in language, it is important, albeit difficult
sometimes, to separate these two perspectives.
The first to deal with sexist lexicography in Spanish was Hampares (1976).
She analysed 105 commonly used job denominations in the Royal Academy’s
dictionary and in two Spanish-English dictionaries, classifying them according
to lower/higher ranked jobs and suffixes (-o vs. -a). Her conclusion was that for
higher ranked job titles only the masculine term was registered (e.g., científico
‘scientist’, diplomático ‘diplomat’), whereas lower ranked jobs did appear in the
feminine form (e.g., asistenta ‘assistant’, sirvienta ‘servant’). Although many
feminine denominations were listed in the dictionary, many of these only
appeared in subsections of the masculine term by giving the explanation ‘wife
of …’ (e.g., embajadora ‘wife of ambassador’). No feminine form was given for

titles like ingeniero ‘engineer’, arquitecto ‘architect’ or mecánico ‘mechanic’.
García Meseguer’s (1977) book also deals with sexist lexicographic practic-
es. The major part of this work is devoted to removing gender stereotypes from
the Academy’s dictionary by listing alternative explanations for words. Howev-
er, when the 20th edition of the dictionary was published in 1984, only a few of
his proposals had been taken into consideration (cf. Forgas Berdet 1986).
The latest edition of the dictionary was published in 1992, but despite slight
improvements in feminine occupational titles, no substantial changes were
introduced (cf. the comprehensive critique of the Academy’s dictionary by
Vargas et al. 1998). Still, numerous shortcomings and sexist practices remain,
for example, about 60 (!) synonyms for the word ‘prostitute’ were considered
worthy and important (?) enough to be included.
Naturally, the Academy has tried to defend itself by arguing that the
dictionary does nothing but register actual language use (cf. Salvador 1990).
However, the crucial question in this matter is not dealt with, namely where the
data that is to reflect actual language use is taken from; women’s magazines or
even feminist publications can hardly be said to be the Academy’s primary
sources of reference.
Although the sexist lexicographical practices of the Academy are particular-
ly obvious, it has to be borne in mind that other authoritative dictionaries do
not necessarily fare much better. Thus, for example, in the two-volume dictio-
nary by María Moliner (1967) the pair mujer and hombre are presented highly
asymmetrically (see Bierbach 1989, 1992: 289) and the explanations of words
given sometimes resemble what García Meseguer had denounced in the
Academy’s dictionary:
(7) Paternidad: Circunstancia de ser padre.
Maternidad: Circunstancia de ser madre: “Le prueba bien la
maternidad”.
‘Paternity: Circumstance of being a father.’
‘Maternity: Circumstance of being mother: “Maternity
is good for her”.’
(Moliner 1967: 668; 366)
(8) Padrino: Hombre que …
Madrina: Femenino de “padrino”. Mujer que …
‘Godfather: Man who …’
‘Godmother: Feminine of “godfather”. Woman who …’
(Moliner 1967: 602; 303)
Spanish 267
While (7) reflects a stereotypical attitude as to the gender roles of women and
men, in (8), the word madrina is presented as if it were derived from padrino,
which, morphologically, is absurd. Even if it were not, the corresponding
explanation for padrino, namely masculino de “madrina”, would still be missing.
Apart from the Academy’s dictionary, we know little about how sexist
recently published mono- and bilingual dictionaries are. The study by Ham-
pares (1976) dates back more than 20 years and more contemporary research
would be welcome.
Contrary to what has been the case for many other languages, investigations
that are based on authentic verbal behaviour of women and men are still very
rare in Spanish,20 and most of the literature published in Spain on this topic
draws primarily on results found for English (cf. e.g., Lozano Domingo 1995).
However, the authors of the book Gramática femenina21 (López García &
Morant 1991) have collected numerous examples (mainly from modern
literature and magazines) that give an impression of the morphological and
lexical features that stereotypically are related to either women or men.22 With
respect to the use of interjections, the authors claim that e.g., ¡huy! often
followed by ¡por Diós! is typical of women, as is ¡ya! As for obscene interjections,
it was at one time assumed that women did not use them. Thus, vulgar interjec-
tions like ¡coño! ‘cunt’ or ¡leche! ‘sperm’, lit. ‘milk’ were reserved for men, while
women “softened” these expressions by modifying the vowel or the consonant:
¡coñe! and ¡leñe! Today, however, younger women in particular do not shrink
from using these or similar expressions, including feminised versions of ¡estoy
hasta los huevos! ‘I am fed up with it [up to my balls! (lit. eggs)]’ as ¡estoy hasta
los ovarios! ‘… up to my ovaries’ and ¡estoy hasta el coño! ‘… up to my cunt’ (cf.
Section 3.3).
If religious dimensions are considered, men of the older generation were
inclined to use blasphemous expressions, like ¡hostia puta! ‘damn it!’, lit.
‘consecrated wafer’ and ‘whore’, or ¡me cago en Dios! ‘I shit on God!’, whereas
women usually made invocations to divine entities: ¡bendita sea Dios! ‘God be
blessed!’ or ¡Virgen Santa! ‘Holy virgin!’. Women and men also differ in their
use of forms of address. Expressions used by women to other women are
typically cariño ‘love’, vida ‘life’ and corazón ‘heart’; men to men use, amongst
others, tronco ‘trunk’, gachó ‘bloke, guy’ and jefe ‘chief ’. A quite peculiar feature
is the fact that the word hombre ‘man’ may be used as an address form for both
men and women (e.g., ¡Qué sí hombre! ‘Certainly I am right, man!’), whereas
mujer ‘woman’ is used only with female interlocutors.
Word-formation patterns of female and male speakers seem to differ.
Women tend to use prefixes like super- or hiper- more frequently than men
(e.g., super-enamorada ‘super in love’, hiper-excitado ‘hyper-excited’), and the
same seems true for diminutive forms. Thus, a female speaker presented a
weather forecast in this way: Hará calor-cillo. Seguiremos con rop-ita de verano.
(‘There will be a bit of warmth. We will go on wearing our nice, tiny summer
clothes’). Men, by contrast, tend to use prefixes like re- or so- to intensify
choleric interjections (as in e.g., ¡Rediós, estáte quieto! ‘For [many] God’s sake,
be quiet!’) or suffixes like -men or -amen, especially in relation to women’s and
men’s erogenous zones (e.g., Tu cul-amen me electriza! ‘Your [nice, lovely]
bottom electrifies me!’). Although the authors cautiously speak of “tendencies”
only, and there is little doubt that the features presented are recognised as
differences due to the speaker’s gender, the data given still lacks empirical
support to determine whether women and men – in comparable situations –
really behave in the way others presumably perceive them to do.
With regard to male-female communicative styles in authentic conversa-
tions, most studies until now have focused on “Anglo-American and Germanic
speakers and languages, that is on ‘Northern’ (post-)industrial societies and
their typical communication settings” (Bierbach 1997a: 107) and Spanish has
not received much attention. In fact, I know of only three investigations that
specifically compare the communicative styles of women and men in Spanish
(Bierbach 1997a, 1997b; Kähler 1997).23
The investigation by Bierbach (1997a) presents material from natural
conversational interaction between working-class people in a neighbourhood
organisation in Barcelona. By analysing the traditional conversational features
of verbal interaction such as interruption, overlap, turn taking, etc., this
investigation yields some surprising results. Contrary to what has been shown
for other languages, the three female informants together talk more than the
five men in the mixed interactions examined. In addition, the women do better
than the men when it comes to self-selected turn-taking: In fact, almost twice as
many women take the floor by themselves.
One feature of Spanish conversations that appears to be difficult to analyse
in terms of “dominance” are interruptions, as these apparently follow different
patterns than those found in non-Mediterranean settings (cf. Bierbach
1997a: 119–122). For example, in Peninsular Spanish, overlaps are quite
Spanish 269
frequent and considered normal: “Nobody ever complains of being interrupted

or uses explicit floor defending formulae” (Bierbach 1997a:121f). Characteristi-
cally, “middle-aged women, especially in lower (working) class milieus in Spain,
behave quite dominantly and self-assured, speaking in a loud voice, joking,
commanding, swearing, etc.” (Bierbach 1997a: 126).
As the overall results of this investigation run contrary to similar findings
in other countries, Bierbach concludes that speakers of Spanish, indeed, behave
differently. However, it should be remembered that the social groups typically
used in other countries for similar experiments consist of people from the
middle class. Therefore, the question asked in the title of Bierbach’s article (“Is
Spain different?”), cannot be answered satisfactorily until data from corre-
sponding “lower” classes in other countries are provided and analysed. The
study by Kähler (1997) is based upon conversational interaction recorded from
Spanish TV-debates, which has been analysed along parameters such as
participation, moderator’s gender, etc. Contrary to Bierbach’s findings, Kähler’s
results confirm, more or less, what has been found in similar studies in other
countries: that men are quantitatively very dominant in mixed discussions, i.e.
they have much longer turns, are more often called upon by the moderator as
well as by the other participants and, in general, are treated as experts more
often than women.
Interestingly, the investigations by Bierbach and Kähler reach different,
almost opposite, conclusions, which suggests that general or even universal
claims about women’s and men’s conversational style should be made with
great caution, if at all. Certainly, Bierbach (1997a) has convincingly demon-
strated that social group membership and situational settings are parameters
which cannot be dispensed with.
5. Guidelines for a non-sexist use of Spanish
The first official guidelines for non-sexist language use – published by the
Spanish Ministry of Education and Science – did not appear until 1988 (Minis-
terio de Educación y Ciencia 1988). Soon afterwards, similar official publica-
tions followed, and today a number of regional institutions in Spain have
published their own guidelines. To the best of my knowledge, no guidelines
have been published in Latin America, but a Spanish version of the UNESCO
guidelines exists (UNESCO 1990).
In general, and apart from some minor differences, all recommendations

propose the same procedures:
a. Propagation of the feminisation of words that refer specifically to women.
b. Replacement of the masculine form for generic reference by epicene words
or by splitted constructions.
c. Avoidance of certain asymmetrical words; e.g., señorita ‘Miss’ should be
replaced by señora24 and hombre ‘man’ by humanidad ‘humanity’ or ser
humano ‘human being’.
Although the guidelines clearly promote the feminisation of occupational titles,

one has to bear in mind that their use and acceptance by speakers of Spanish
vary considerably according to the title in question. A questionnaire investiga-
tion from 1991 conducted in Spain revealed, e.g., that almost half of the
respondents preferred primera ministra to primer ministro, but only 17%
preferred the (controversial) form la jueza ‘the female judge’ to la juez, a noun
of double gender (Nissen 1991: 180–182). On the other hand, attitudes towards
specific job titles are changing rapidly, and it is very likely that results would be
different today as there are, indeed, more female judges in Spain than male
ones. Phonetic criteria and analogies to similar words also exert constraints on
the acceptance of specific expressions.
A further aspect addressed by some of the guidelines is the order of words
that refer to men and women. Except for certain direct forms of address (e.g.,
Damas y caballeros ‘Ladies and gentlemen’), the convention is to put the
feminine after the masculine form. As the authors of the guidelines interpret
this traditional word order as a reflection of the hierarchical (or patriarchal)
structure of society, they recommend an inversion or variation of order, e.g., las
mujeres y los hombres ‘women and men’, las profesoras y los profesores ‘female
and male teachers’.
Despite their good intentions, the guidelines themselves continue to use the
ambiguous masculine form as a generic, thereby revealing the difficulties the
proposed changes may imply:
(9) … sin dar por sentado que todas las mujeres tienen como objetivo único el
matrimonio y los hijos (m).
‘… without assuming that the only aim of all women is marriage and
children (m).’
(Ministerio de Educación y Ciencia 1988: 11)
Spanish 271
Traditional practice may also explain why the usual Varios autores (m) ‘Various
authors’ is used in the bibliographical section of a co-authored book instead of
Varias autoras (f) even though the book concerned has only female contributors
(Lasaosa Zazu 1994: 16).25
Today, more than one decade after the first guidelines were published, it is
still too early to say what impact these and similar publications have had on the
Spanish language. Although words that refer to women in the vast majority of
texts have been feminised, the substitution of masculine words used generically
with alternative expressions is infrequent in ordinary texts.
Nonetheless, the guidelines have undoubtedly promoted the use of femi-
nine forms for female referents and aided the process of making these words
part of dictionaries and other language reference material. The consistent
avoidance of the ambiguous masculine form can also be noted in public
administration when official forms, applications, letters to clients, etc. are used,
but apart from feminist journals and (some) publications by the Instituto de la
Mujer, the masculine form prevails exuberantly for generic references in
ordinary texts, such as books, newspapers, magazines, etc. Even many (female)
authors of books which specifically address “language and gender” issues
continue to use the masculine forms, often briefly commenting on their reasons
for doing so in the introduction (cf., e.g., Lozano Domingo 1995).
As mentioned, in certain types of texts (e.g., official forms, documents, job
advertisements), the masculine form is avoided and replaced by splitted
expressions. In these texts, a variety of forms abbreviated and segmented by
means of brackets, slashes or hyphens are used, e.g., obrero(a); obrero/a; obrero-
a ‘worker’, although, paradoxically, none of these were ever mentioned in any
of the published guidelines. Another phenomenon which indicates that
language users can surpass the creativity of guidelines is a new version of the
splitting procedure: Probably in accordance with the controversial so-called
Binnen-I in German (as in LehrerInnen ‘female or male teachers’, cf. Bußmann
& Hellinger, vol. III), some users of Spanish have graphically combined the
word endings -a and -o to form l@s alumn@s ‘the pupils’ (Arias Barredo
1995: 62). Being the most economical alternative, it is likely to spread rapidly
(cf. also the critical analysis by Porto Dapena 1999).
Predictably, immediately after the publication of the guidelines, there were
reactions against them. On the whole, feminisation of occupational titles was
generally accepted for specific reference. However, when titles refer to a
profession in general (as is the case on university diplomas), the administrations
of the institutions involved showed a remarkable reluctance to employ the

feminine terms – an issue which frequently hit the newspaper headlines.
The main target of criticism was the procedure of splitting in order to avoid
the masculine form. In the majority of cases, the critics managed to imagine
contexts in which the splitting procedures appeared convincingly superfluous.
For example, should people in an emergency situation really call for una médica
o un médico ‘a female or a male doctor’ instead of just un médico ‘a doctor’
(López García & Morant 1991: 56), or should the guests at a wedding party
shout ¡Viva la novia y el novio! (‘Long live the bride and the bridegroom!’)
instead of ¡Vivan los novios (m)! ‘Long live the couple!’ (Lázaro Carreter
1988: 10) and thus change a century old tradition?26 One member of the
Academy claimed that it would be “superfluous and even ridiculous” to recast
the phrase Me acuerdo mucho de mis padres (m) ‘I often think of my parents’
into Me acuerdo mucho de mi padre y de mi madre ‘I often think of my father
and of my mother’ (Alarcos Llorach 1990: 136). Naturally, many situational
contexts will unmistakably disambiguate the masculine form (as e.g., in the
latter case because nobody is likely to have two fathers).27
6. Conclusion
The first book on language and gender was published in Spain more than twenty
years ago (García Meseguer 1977), and one may be surprised at how much and
how little has been achieved since. On the one hand, the wording in the Royal
Academy’s dictionary continues to be a target for criticism by those who
denounce sexist practices, but, as even today there is only one female member of
the Academy, it is doubtful that radical changes will be made in the near future.
On the other hand, and perhaps even due to the blatant conservatism of the
Academy, language and gender matters have become commonplace in the
newspapers, in letters to the editor and even, sometimes, on TV. For example,
in March 1994, Televisión Española broadcast a chat show dedicated to the theme
of La Mujer y el Hombre en el Lenguaje ‘Woman and man in the language’.
As for linguistic reform, new guidelines for non-sexist language are being
published even by local women’s institutes in the various regional autonomies
in Spain. However, traditional language planning, which is associated with
academia, is, more often than not, confronted with hostility and with counter-
arguments to positions that none of the guidelines have ever claimed explicitly
(e.g., that language reform can change society).
Spanish 273
Despite this attitude towards official guidelines, publications about gender

issues in Spanish within the last ten years have, at least, tripled in comparison
to the decade before. However, many publications are not necessarily the fruit
of academic labours in Spain (nor in other Spanish-speaking countries), but are
frequently due to foreign language departments in other countries. Research on
the discourse behaviour of women and men is still at its very beginnings in
Spain and many recently published works on e.g., colloquial Spanish have had
to draw heavily on publications that are more than 30 years old. Conversational
analyses with a special focus on gender are practically non-existent in Spanish.
Similarly, there is a dearth of cross-linguistic and contrastive gender studies.
Although numerous varieties of Spanish are spoken in the world, studies that
compare gender issues in different countries are totally absent. Thus, it would
be desirable, for example, to compare Spain and Mexico (the country that is
often regarded as the stereotype with respect to machismo), in terms of male and
female communicative behaviour, as more or less the same language is spoken
in both countries.
Notes
1. For a general introduction to Spanish, see Asher (1994) or Bright (1992); useful grammar
books are Butt & Benjamin (2000) and de Bruyne (1995); for contrastive studies see Whitley
(1986). The most comprehensive monolingual dictionary is Moliner (1967). For regional
varieties of Latin American Spanish see Lipski (1994); for Peninsular Spanish see Zamora
Vicente (1970).
2. Due to limitations of space, communicative behaviour in bilingual settings is not treated
here. For references on this matter see the introductory articles by Bierbach (1992: 281–281)
and Nissen (1990: 19–22). Also, studies that treat phonological differences are not dealt with,
but see the overview articles by Bierbach (1992:279–281), Kowallik (1997), Lozano Domingo
(1995: 82–122), Nissen (1990: 17–18) or Rissel (1981), as well as the recent empirical studies
by Almeida (1995), Terborg Schmidt (1992) and Woolard (1997). For references on proverbs
and sexism see the comprehensive work by Calero Fernández (1991). For references on el
piropo i.e. ‘a flattering/insulting expression used by male speaker(s) to address women’, see
Bierbach (1992: 282–283), and López García & Morant (1991: 170–173), as well as the
compilation by Diosdado (1990). For nonverbal communication, see López García & Morant
(1991: 65–88).
3. All translations are my own.
4. It should be mentioned that a consistent terminology, especially in relation to contrastive
studies, is much needed. One example of this need is reflected in Fuertes Olivera’s (1992)
inspiring book about gender stereotypes in English and Spanish. He suggests substituting
común with dual ‘double’ (Fuertes Olivera 1992: 51, footnote 6), but – apart from not
following his own proposal as he continues to use común on the subsequent pages – he fails
to notice the difference between a noun of double gender and an epicene noun (see
Section 2.2) by classifying the word profesorado ‘staff of teachers’ as double (Fuertes Olivera
1992: 165). Although both an epicene and a double gender noun may refer to both women
and men, an epicene noun has one and only one gender (profesorado, m) and cannot,
therefore, be a noun of double gender. Nouns of double gender, on the other hand, are
always either masculine or feminine depending on the referent.
5. For a detailed categorisation see García Meseguer (1991).
6. Most animals do not have separate words for females and males. If necessary they are
distinguished by adding hembra ‘female’ or macho ‘male’ to the noun.
7. For other examples see the appendix in Fuertes Olivera (1992), in which he presents
more than 400 potential neologisms (i.e. words not registered in the dictionary of the
Academy of 1984) that refer to human beings. All neologisms were formed according to the
pattern non -a vs. -a.
8. García Meseguer (1977, 1984) also uses the same graphic procedure, when he proposes
alternative, non-sexist formulations.
9. Surprisingly, the graphic representation of words that add -a is identical, e.g., japonés, -a
(‘Japanese, male/female’). In fact, only knowledge of Spanish word formation in general
prevents the production of non existing words like niñoa (from niño,-a, formed in analogy
to e.g., japonesa).
10. However, for jobs that traditionally, and still today, are done by women, the masculine
term does not automatically function as a generic. Thus, Los enfermeros (m) están de huelga
‘The nurses are on strike’ will for most speakers probably not include women. On the other
hand, the feminine title will not be used for generic reference either; instead, personal de
enfermería (‘staff of the infirmary’) is employed (cf. Fernández Lagunilla 1991: 326).
11. The term neutral here stresses the semantic aspect and differs, therefore, from gender-
neutral as used by Frank & Treichler (1989: 18): “a gender-neutral term is formally, linguisti-
cally unmarked for gender”.
12. Interestingly, the word form persono exists in Esperanto (Wolfe 1980:789), where it is not
a masculine but a genderless term.
13. Note that with respect to animals this is not true. For example, abeja (f) ‘bee’ and gallina
(f) ‘hen’ are the generic terms for abeja/zángano and for gallina/gallo; cf. carne de gallina
(*carne de gallo) ‘goose pimples’, gallinero (*gallero) ‘hen run’. Similarly, zángano is usually
defined on the basis of abeja, i.e. abeja macho ‘male bee’.
14. For a more detailed discussion of this matter and also on how agreement problems are
solved in splitted constructions see Nissen (1997b). For other problems of agreement, see
e.g., Wonder (1985).
15. A similar investigation for German showed that one sentence that happened to contain
the masculine pronoun er ‘he’ scored the highest rating of “male bias” (Klein 1988: 315).
16. For further reading on occupational titles see García Meseguer (1993) and DeMello (1990).
Spanish 275
17. More information about sexual terminology can be obtained from Bierbach
(1992: 289–291), Suardiaz (1973), and López García & Morant (1991: 139–156). Empirical
studies, however, that deal with real language use are lacking. See also Castro (1982).
18. It should be noted that some earlier investigations on this matter had been carried out a
few years previously in Puerto Rico, undoubtedly under the influence of gender discussions
in the US. The results obtained agree almost entirely with those found later on in Spanish
schoolbooks (cf. González 1985: 49–50).
19. For further references see Careaga Castrillo & Garreta Torner (1987), Moreno (1986),
Catalá Gonzalvez & García Pascual (1987), and Moreno Marimón (1994).
20. Exceptions are investigations that deal with purely phonological aspects or analyse the
behaviour in bilingual settings, cf. note 2.
21. The title of the book is misleading, as no traditional grammatical aspects are addressed.
Instead, the verbal (and nonverbal) behaviour of women and men is treated. Even if one
considers these aspects as part of grammar, there still seems to be no reason to call the book
Gramática femenina as both female and male behaviour is covered. The following examples
are all from the second chapter of this book.
22. Although the line of demarcation can be somewhat arbitrary, I have left out those
sections that do not specifically address verbal differences, but rather describe specific
situations that are typical for women or men. For example, a sentence like ‘I cannot go. I do
not know what to put on.’ (López García & Morant 1991: 124) may be said more often by
women, but is it, therefore, a characteristic feature of women’s language?
23. Another discourse study related to language and gender is Martín Rojo & Callejo Gallego
(1995). However, the aim of the study is not to compare men’s and women’s conversational
styles, but, instead, to show how men (in this case Spanish executives) speak about women,
and how they implicitly favour discrimination in recruitment and promotion.
24. Although all guidelines recommend the avoidance of señorita ‘Miss’, there is no
agreement as to the abbreviation of its substitute, i.e. señora. The Ministry for Social Affairs
has tried to coin a new expression: Sa. (Ministerio de Asuntos Sociales 1991: 9), equivalent
to the English ‘Ms’, while the UNESCO (1990: 14) prefers Sra. because it is known through-
out the world – despite the fact that it traditionally only referred to a married woman.
25. The book in question is Moreno Marimón (1994).
26. Although there are no other possibilities in the Spanish-speaking countries, in Denmark
the word novios (m) could refer to a gay male couple, since homosexuals are allowed to
marry legally (although without approval of the Church).
27. Nevertheless, Alarcos Llorach’s example is not well chosen, as it differs from “ordinary”
splitted expressions (e.g. obrera y obrero ‘worker’). In this case, two different lexemes are
involved (madre and padre) and the construction may not, therefore, be felt to be as
“awkward” as many splitted expressions in fact are. Futhermore, most people probably
regard their parents as individual persons anyway and, therefore, an explicit mentioning of
both persons may not be perceived as superfluous at all.
References
Alarcos Llorach, Emilio. 1990. “Género, número y sexo.” ABC, Feb. 2, 1990: 36.
Almeida, Manuel. 1995. “El factor ‘sexo’ en los procesos de variación y cambio.” Anuario de
Letras 33: 97–109.
Arias Barredo, Aníbal. 1995. De feminismo, machismo y género gramatical. Valladolid:
Universidad de Valladolid.
Asher, Ronald E. ed. 1994. The encyclopedia of language and linguistics. 10 vols. Oxford:
Pergamon.
Berro-García, Adolfo. 1952/53. “Formación del femenino en los nombres de profesiones,
oficios y actividades ejercidos por mujeres.” Boletín de Filología 7: 510–514.
Bierbach, Christine. 1989. “La lengua, compañera del imperio macho. Was ist sexistischer:
die spanische Sprache oder ihre Wörter?” Tranvía 14: 7–9.
Bierbach, Christine. 1992. “Spanisch: Sprache und Geschlechter.” In Lexikon der Romanis-
tischen Linguistik, eds. Günter Holtus & Michael Metzeltin & Christian Schmitt.
Tübingen: Niemeyer, 276–296.
Bierbach, Christine. 1997a. “Is Spain different? Observations on male-female communicative
styles in Spanish group discussion.” In Communicating gender in context, eds. Helga
Kotthoff & Ruth Wodak. Amsterdam: Benjamins, 107–138.
Bierbach, Christine. 1997b. “Starke Frauen! Interaktionsmuster und Gesprächsstrategien in
spanischen Cross-gender-Diskussionen.” In Sprache und Geschlecht in der Romania.
Romanistisches Kolloquium X, eds. Wolfgang Dahmen & Günter Holtus & Johannes
Kramer & Michael Metzeltin & Wolfgang Schweickard & Otto Winkelmann. Tübingen:
Narr, 287–320.
Bright, William, ed. 1992. International encyclopedia of linguistics. 4 vols. Oxford: Oxford
University Press.
Bustamente y Rivero, José Luis. 1960. “Doble terminación para oficios y profesiones.” In
Tercer Congreso de Academias de la Lengua Española, ed. Academia Colombiana de la
Lengua. Bogotá: Academia Colombiana de la Lengua, 421–424.
Butt, John & Carmen Benjamin. 2000. A new reference grammar of modern Spanish. 3rd ed.
London: Arnold.
Calero Fernández, María Ángeles. 1991. La imagen de la mujer a través de la tradición
paremiológica española (lengua y cultura). 2 vols. Barcelona: Universitat de Barcelona.
Calero Fernández, María Ángeles. 1999. Sexismo lingüístico. Análisis y propuestas ante la
discriminación sexual en el lenguaje. Madrid: Narcea.
Careaga Castrillo, Pilar & Nuria Garreta Torner. 1987. Modelos masculino y femenino en los
textos de EGB. Madrid: Ministerio de Cultura. Instituto de la Mujer.
Casares, Julio. 1947. Divertimientos filológicos. Obras completas. 2nd ed. Madrid: Espasa
Calpe, vol. III: 302–306.
Castro, Rafaela. 1982. “Mexican women’s sexual jokes.” Aztlán 13: 275–293.
Catalá Gonzalvez, Aguas Vivas & Enriqueta García Pascual. 1987. Una mirada otra. Valencia:
Generalitat Valenciana, Departamento de la Dona.
Spanish 277
de Bruyne, Jacques. 1995. A comprehensive Spanish grammar. Adapted with additional

material by Christopher J. Pountain. Oxford: Blackwell.
DeMello, George. 1990. “Denotation of female sex in Spanish occupational nouns.”
Hispania 73: 392–400.
Diosdado, Esteban. 1990. Los mejores piropos. Barcelona: Edicomunicación.
Errazu Colás, María Ángeles. 1995. Hacia una utilización no sexista del lenguaje. Zaragoza:
Instituto Aragonés de la Mujer.
Fernández de la Torre Madueño, María Dolores, Antonia María Medina Guerra & Lidia
Taillefer de Haya. 2000. El sexismo en el lenguaje: actas del congreso sobre el sexismo en el
lenguaje celebrado los días 4, 5, 6 de noviembre de 1998 en la UMA. 2 vols. Málaga:
Diputación Provincial de Málaga. Servicio de Publicaciones.
Fernández Lagunilla, Marina. 1991. “Género y sexo: controversia científica o diálogo de
sordos?” In Actas de las VIII Jornadas de investigación interdisciplinaria. Los estudios sobre
la mujer: de la investigación a la docencia, eds. Cristina Bernis & Violete Demonte & Elisa
Garrido & Teresa G. Calbet & Isabel de la Torre. Madrid: Instituto Universitario de
Estudios de la Mujer, Universidad Autónoma de Madrid, 319–328.
Forgas Berdet, Esther. 1986. “Sexo y sociedad en el último DRAE.” Universitas Tarra-
conensis 10: 79–100.
Frank, Francine Wattman & Paula Treichler. 1989. Language, gender and professional writing.
New York: Modern Language Association.
Fuertes Olivera, Pedro A. 1992. Mujer, lenguaje y sociedad: Los estereotipos de género en inglés
y en español. Valladolid: Universidad de Valladolid.
García Meseguer, Álvaro. 1977. Lenguaje y discriminación sexual. Madrid: Cuadernos para el
diálogo.
García Meseguer, Álvaro. 1984. Lenguaje y discriminación sexual. 2nd ed. Barcelona: Monte-
sinos.
García Meseguer, Álvaro. 1991. “Gender-sex clashes in Spanish: A semantic typology of
animated nouns.” Journal of Pragmatics 15: 445–463.
García Meseguer, Álvaro. 1993. “Género y sexo en el nuevo Diccionario de la Real Acade-
mia.” Política científica 37: 51–56.
García Mouton, Pilar. 1999. Cómo hablan las mujeres. Madrid: Arco/Libros.
González, Iris G. 1985. “Some aspects of linguistic sexism in Spanish.” In Sprachwandel und
feministische Sprachpolitik: Internationale Perspektiven, ed. Marlis Hellinger. Opladen:
Westdeutscher Verlag, 48–63.
Hampares, Katherine. 1976. “Sexism in Spanish lexicography?” Hispania 59: 100–109.
Hellinger, Marlis. 1990. Kontrastive feministische Linguistik. Mechanischen sprachlicher
Diskriminierung im Englischen und Deutschen. München: Hueber.
Kähler, Gisela. 1997. “Geschlechtsspezifische Sprechweisen in spanischen Fernseh-
diskussionen.” In Sprache und Geschlecht in der Romania. Romanistisches Kolloquium X,
eds. Wolfgang Dahmen & Günter Holtus & Johannes Kramer & Michael Metzeltin &
Wolfgang Schweickard & Otto Winkelmann. Tübingen: Narr, 261–270.
Klein, Josef. 1988. “Benachteiligung der Frau im generischen Maskulinum – eine feminis-
tische Schimäre oder psycholinguistische Realität?” In Das Selbstverständnis der
Germanistik, ed. Norbert Oellers. Tübingen: Niemeyer, 310–319.
Kowallik, Sabine. 1997. “Zur Erfassung geschlechtsspezifischer Charakteristika der spanischen

Sprache.” In Sprache und Geschlecht in der Romania. Romanistisches Kolloquium X, eds.
Wolfgang Dahmen & Günter Holtus & Johannes Kramer & Michael Metzeltin &
Wolfgang Schweickard & Otto Winkelmann. Tübingen: Narr, 271–286.
Lasaosa Zazu, Asunción. 1994. La comunicación como elemento de progreso. Navarra:
Gobierno de Navarra, Instituto Navarro de Bienestar Social, Subdirección de la Mujer.
Lázaro Carreter, Fernando. 1988. “Ministro, ministra, ministros.” ABC, suplemento, July 31,
1988: 10.
Lebsanft, Franz. 1997. Spanische Sprachkultur. Tübingen: Niemeyer.
Lipski, John. 1994. Latin American Spanish. London: Longman.
Lledó, Eulàlià & Mercè Otero. 1994. “El sexismo en la lengua y en la literatura.” In Del
silencio a la palabra, ed. Montserrat Moreno Marimón. Madrid: Ministerio de Asuntos
Sociales, Instituto de la Mujer, 358–379.
López García, Ángel & Ricardo Morant. 1991. Gramática femenina. Madrid: Cátedra.
Lozano Domingo, Irene. 1995. Lenguaje femenino, lenguaje masculino: condiciona nuestro sexo
la forma de hablar? Madrid: Minerva.
Lyons, John. 1977. Semantics. 2 vols. Cambridge: Cambridge University Press.
Martín Rojo, Luisa & Javier Callejo Gallego. 1995. “Argumentation and inhibition: Sexism
in the discourse of Spanish executives.” Pragmatics 5: 455–484.
Ministerio de Asuntos Sociales. 1991. Uso no sexista del lenguaje administrativo. Madrid:
Ministerio de Asuntos Sociales.
Ministerio de Educación y Ciencia. 1988. Recomendaciones para un uso no sexista de la
lengua. Madrid: Ministerio de de Educación y Ciencia.
Moliner, María. 1967. Diccionario de uso del español. 2 vols. Madrid: Gredos.
Moreno, Montserrat. 1986. Como se enseña a ser niña: El sexismo en la escuela. Barcelona: Icaria.
Moreno Marimón, Montserrat, ed. 1994. Del silencio a la palabra. Madrid: Ministerio de
Asuntos Sociales, Instituto de la Mujer.
Müller, Bodo. 1987–2000. Diccionario del español medieval. 5 vols. Heidelberg: Winter.
Nebrija, Antonio. 1492 [1990]. Gramática de la lengua castallana, ed. Antonio Quilis.
Madrid: Nacional.
Nissen, Uwe Kjær. 1986. “Sex and gender specification in Spanish.” Journal of Pragmatics 10:
725–738.
Nissen, Uwe Kjær. 1990. “A review of research on language and sex in the Spanish language.”
Women and Language 13: 11–29.
Nissen, Uwe Kjær. 1991. Feminiseringstendenser i moderne spansk [Tendencies towards the
feminisation of modern Spanish]. Ph.D. dissertation. Odense University, Odense,
Denmark.
Nissen, Uwe Kjær. 1997a. “Do sex-neutral and sex-specific nouns exist: The way to non-
sexist Spanish?” In Kommunikation von Geschlecht. Communication of gender, eds.
Friederike Braun & Ursula Pasero. Pfaffenweiler: Centaurus, 222–241.
Nissen, Uwe Kjær. 1997b. “Soziolinguistik der Syntax? Genuskongruenz im Spanischen aus
feministischer Sicht.” In Sprache und Geschlecht in der Romania. Romanistisches
Kolloquium X, eds. Wolfgang Dahmen & Günter Holtus & Johannes Kramer & Michael
Metzeltin & Wolfgang Schweickard & Otto Winkelmann. Tübingen: Narr, 321–344.
</TARGET "nis">
Spanish 279
Olivares, Carmen. 1984. “A comment on J. L. Mey’s review article ‘Sex and language
revisited’.” Journal of Pragmatics 8: 753–756.
Olivares, Carmen. 1986. “A reply to Nissen.” Journal of Pragmatics 10: 739–740.
Perissinotto, Giorgio. 1983. “Spanish ‘hombre’: Generic or specific?” Hispania 66: 581–592.
Perissinotto, Giorgio. 1985. “La producción de géneros específicos en contextos condiciona-
dos.” Estudios de Lingüística Aplicada 3: 112–128.
Porto Dapena, José Álvaro. 1999. “Género ‘arroba’, neutralización masculino/femenino y
síndrome antimachista.” Español Actual 72: 5–14.
Pusch, Luise F. 1980. “Die männliche Gruppe als referenzsemantische Grundeinheit.”
Linguistische Arbeiten und Berichte 15: 178–181.
Real Academia Española. 1970. Diccionario de la lengua española. 19th ed. Madrid: Espasa-
Calpe. 20th ed. (1984), 21st ed. (1992).
Real Academia Española. 1977. Esbozo de una nueva gramática de la lengua española. Madrid:
Espasa-Calpe.
Rissel, Dorothy. 1981. “Diferencias entre el habla femenina y la masculina en español.”
Thesaurus 36: 305–322.
Salvador, Gregorio. 1990. “El diccionario y la gente.” In Homenaje al Profesor Francisco
Marsá. Jornadas de filología, Universitat de Barcelona. Barcelona: Universitat de
Barcelona, 193–207.
Spender, Dale. 1980. Man made language. London: Routledge.
Suardiaz, Delia Esther. 1973. Sexism in the Spanish language. Studies in Linguistics and
Language Learning IX. MA thesis. University of Washington, Seattle.
Terborg Schmidt, Roland. 1992. “El papel de la mujer en el cambio lingüístico. Una
propuesta para la clasificación de situaciones de conflicto.” Estudios de Lingüística
Aplicada 10: 242–253.
Teschner, Richard & William M. Russel. 1984. “The gender patterns of Spanish nouns: An
inverse dictionary-based analysis.” Hispanic Linguistics 1: 115–132.
UNESCO. 1990. Recomendaciones para un uso no sexista del lenguaje. Paris: UNESCO.
Vargas, Ana & Eulàlia Lledó & Mercedes Bengoechea & Mercedes Mediavilla & Isabel Rubio
& Aurora Marco & Carmen Alario. 1998. Lo femenino y lo masculino en el Diccionario de
la Lengua de la Real Academia Española. Madrid: Ministerio de Trabajo y Asuntos
Sociales. Instituto de la Mujer.
Whitley, M. Stanley. 1986. Spanish-English contrasts: An introduction to Spanish linguistics.
Georgetown University Press.
Wolfe, Susan J. 1980. “Gender and agency in Indo-European languages.” Papers in Linguis-
tics 13: 773–794.
Wonder, John P. 1985. “Género natural, género gramatical.” Hispania 68: 273–282.
Woolard, Kathryn A. 1997. “Between friends: Gender, peer group structure, and bilingualism
in urban Catalonia.” Language in Society 26: 533–560.
Zamora Vicente, Alonso. 1970. Dialectología española. Madrid: Gredos.
<LINK "pha-n*">
<TARGET "pha" DOCINFO
AUTHOR "Hoa Pham"
TITLE "Gender in addressing and self-reference in Vietnamese"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
vietnamese
Gender in addressing and self-reference

in Vietnamese
Variation and change*
Hoa Pham
University of Toronto, Canada
1. Introduction
2. Selected structural properties of Vietnamese
3. Terms for human referents
3.1 Personal nouns
3.2 Kinship terms
3.3 Pronouns
4. Male/female relations in idiomatic expressions
5. Vietnamese terms of address, reference and self-reference
5.1 Sociopolitical background and characteristics of address terms
5.2 Kinship terms and common nouns
5.3 Terms of address and self-reference in various female-male encoun-
ters
5.3.1 Masters and housekeepers
5.3.2 Teachers and students
5.3.3 Neighbors and work mates
5.4 Women and men as couples
6. Conclusion
Notes
References
1. Introduction
Vietnamese, called Annam under Chinese domination, today known to its

native speakers as tieâ ng Viêt
» nam, is the official language of Vietnam and
282 Hoa Pham
belongs to the Viet-Muong branch of the Mon-Khmer subfamily of Austro-

Asiatic languages. It is spoken by about 76 million people (Dodd & Lewis 1998).
The Northern standard language is based on the Hanoi dialect. The Southern
standard dialect is based on the speech of people living in Saigon (now Ho Chi
Minh City). There is also a Central standard dialect focused on the speech of
Hue. The Hanoi dialect is closest to the writing system, which encodes all those
contrasts that are found in all dialects. The dialects form a continuum from
north to south. With regard to tone, Northern and Southern dialects share
many features, while Central dialects stand on their own. Vietnamese is spoken
not only in Vietnam, but also in parts of Cambodia, Laos and Thailand. After
the political victory of the North in 1975, people fled the country in great
numbers. There are about two million Vietnamese people living all over the
world, half of them in North America.1
In the several centuries of Chinese domination, Vietnam used Chinese
characters (ch«þ Hán) in education and official communication. By the 13th
century, the Vietnamese had created their own writing system called ch«þ Nôm,
based on Chinese characters with Sino-Vietnamese pronunciation (see Ðào
1973: 55). In the 16th century, Alexandre de Rhodes, a French Catholic mission-
ary who was a gifted linguist, invented a new writing system based on Roman
scripts, chþ̃ Quâoc ngþ̃ ‘national script’, for religious texts. This system is
phonological in nature. Since the beginning of the 20th century, it has served as
the official writing system of Vietnamese.
2. Selected structural properties of Vietnamese
A large proportion of the vocabulary of Vietnamese is of Chinese origin as the

result of the intensive Chinese cultural influence during ten centuries of
Chinese rule. These lexical items are usually known as Sino-Vietnamese. Most
Sino-Vietnamese (SV) items have Vietnamese (VN) counterparts, which are
interchangeable in most cases, e.g., 0àn bà (VN) / phu» nþ̃ (SV) ‘woman’.
Except for many Sino-Vietnamese compounds where the usage of Vietnamese
counterparts is considered inappropriate, Vietnamese forms can be used in
almost any context, while Sino-Vietnamese forms occur mainly in formal
contexts and written language. The overuse of Sino-Vietnamese is regarded as
pompous and awkward; the overuse of Vietnamese forms for Sino-Vietnamese
compounds is seen as uneducated and improper. However, there have been
Vietnamese 283
concerted efforts by nationalists to replace Sino-Vietnamese forms with

Vietnamese ones where possible.
Vietnamese does not have grammatical gender. It is a classifier language
which demands the use of a classifier when a noun is combined with a numeral.
The choice of classifier depends on features such as animateness, humanness,
social position and attitude of the speaker, to name a few. For instance, con
‘animal’ is used for animals and insects (con kiâen ‘an ant’), and cái ‘thing’ for
inanimate objects (cái bàn ‘a table’). However, con also can be used for inani-
mate objects to emphasize motion, e.g., con thuyáen ‘a boat’. Objects can have
classifiers according to their shape, e.g., cây for stick-like objects (cây bút tree +
pen ‘a pen’), tǜ for sheet-like objects (tǜ giâay ‘a piece of paper’), tâp
» for books
» vüÑ ‘a notebook’). A person can have different classifiers, including a
(tâp
common noun such as ngþǜi (ngþǜi lính person + soldier ‘a soldier’) or nhà
(nhà thü house + poetry ‘a poet’), a kinship term (anh/chú/ông lính older
brother/uncle/grandfather + soldier ‘a soldier’), depending on the age of the
speaker compared with that of the referent; or classifiers such as thÓang ‘fellow’,
tên ‘name’ exhibiting semantic derogation, e.g., thÓang lính ‘a soldier’, tên cþǘp
name + rob ‘a robber’.
In a noun phrase with a numeral, a classifier must be used between the
number and the noun: numeral-classifier-noun, e.g., ba ngþǜi con three +
person + offspring ‘three children’ (Nguyeà n 1987; see also Löbel 2000 for a
detailed description of the occurrence vs. non-occurrence of classifiers).
Vietnamese is a monosyllabic language. Each syllable usually constitutes a
word. The canonical syllable structures are (C)(w)V(C), in which V, i.e., the
only obligatory element, can be a vowel or a diphthong.2 Any consonant can
occur initially, but only a limited number of consonants can occur finally. Every
syllable has a tone. There are six phonemic tones (their Vietnamese names are
» hÑoi and ngã) to distinguish words with identical consonant
ngang, huyáen, sÔac, năng,
and vowel sequences. In Southern dialects, the tone nga merges with the tone hoi.
Variation of the basic word order SVO is significant in Vietnamese, as
might be expected in the virtual absence of other overt indicators to express the
syntactic relations of words to each other. Vietnamese is morphologically
isolating and makes use of modifiers rather than of affixes to express tense and
voice of verb forms, e.g., nó ăn he/she + to eat ‘he/she eats’, nó 0ã (past mor-
pheme) ăn ‘he/she ate’, nó ăn ráoi (already) ‘he/she has already eaten’, nó chþa
(not yet) ăn ‘he/she has not eaten yet’. There are no case markers in Vietnamese,
and thus a noun can function as subject, object, or any other syntactic role without
changing its form (Nguyen à 1977, Thompson 1965, Ngô 1984).
284 Hoa Pham
3. Terms for human referents
This section is concerned with Vietnamese terms for person reference that are
used in addressing, reference and self-reference. As there is no grammatical
gender in Vietnamese, this is an important area where gender differences are
expressed and constructed. The respective terms can be derived from personal
nouns, common nouns, kinship terms, pronouns or proper nouns. In this
section, I shall describe each subsystem separately, in order to establish what
might be called a protosystem, against which I can then outline the changes that
are taking place.
3.1 Personal nouns

A personal noun can be formed from a verb or adjective with a classifier. The
classifier precedes the noun in Vietnamese forms and follows the noun in Sino-
Vietnamese forms. For instance, Vietnamese forms such as ngþǜi 0oc » person +
read ‘a reader’, ngþǜi bênh
» person + sick ‘a patient’, thay á thuâoc master + medi-
á day
cine ‘a physician’, thay » master + teach ‘a teacher’ have the corresponding
Sino-Vietnamese forms 0ôc » giaÑ ‘a reader’, bênh
» nhân ‘a patient’, bác sı̃ ‘a
physician’, giáo sþ ‘a professor’, respectively.
Some verbs, nouns or adjectives can function as nominal address forms for
the interlocutor, e.g., 0Ñe ‘to give birth’ > ‘mother’, cþng ‘to dearly love’ >
‘sweetheart’, vú ‘breast’ > ‘wet-nurse’, nhoÑ ‘small’ > ‘you’.
Except for lexical gender words such as vua ‘king’ vs. hoàng hâu » ‘queen’,
hoàng tþÑ ‘prince’ vs. công chúa ‘princess’ and certain kinship terms (cf. Sec-
tion 3.2), all human nouns and occupational terms are used for both women
and men. As Vietnamese does not have affixes, gender-specification is achieved
by special morphemes used as modifiers (e.g., Sino-Vietnamese trai/gái,
nam/nþ̃ ‘male/female’) or classifiers, e.g., ban » trai ‘a male friend/boyfriend’ vs.
» gái ‘a female friend/girlfriend’; ông hiêutrþ
ban » üng
Ñ ‘Mr Principal’ vs. bà hiêu-»
trþüng
Ñ ‘Mrs Principal’; câu » giáo uncle + teach ‘a male teacher’ vs. cô giáo young
lady + teach ‘a female teacher’; Sino-Vietnamese nam sinh ‘a male student’ vs.
nþ̃ sinh ‘a female student’.3 These forms are used in normal or formal contexts.
There is no asymmetry in these ‘male’/‘female’ morphemes, but the two
particular Vietnamese special designations, thÓang (male) and con (female) occur
only in solidarity contexts and will be discussed in Section 5.4.
Vietnamese 285
3.2 Kinship terms

The system of kinship terms that may be used as address terms is shown in
Table 1, which represents the Southern dialect. In the table, the slash symbol
marks a male-female couple of the same generation. For example, câu » is
mother’s brother, and his wife is mü.» Three generations of relatives are shown
in the table, separated by double lines.
Table 1.Vietnamese kinship terms

ông bà
grandfather grandmother
bác / bác cô / dþüng

» » / mü»
câu dì / dþüng
»
father’s older brother / father’s sister mother’s brother mother’s sister
his wife / her husband / his wife / her husband
chú / thím
father’s younger brother /
his wife
ba me»
father, dad mother, mom
anh ch»i
older brother older sister
con
offspring (term of self-reference in speaking to older generations)
em
younger brother or sister (term of self-reference in speaking to siblings)
Table 1 shows the terms as they would be if the speaker was a member of
the youngest generation. In South Vietnam, a child can use the term con
‘offspring’ for self-reference when speaking to members of the older genera-
tions.4 In the third generation, younger brothers or sisters address their older
brother and sister as anh or ch»i, respectively, and use the term em ‘younger
brother/sister’ for themselves.
Kinship terms differentiate the grandparents by gender (grandmother and
grandfather), but not by lineage (paternal or maternal). However, children
usually use ông nôi» and bà nôi
» to refer to their paternal grandfather and grand-
mother, and ông ngoai » and bà ngoai» to refer to their maternal grandfather and
286 Hoa Pham
grandmother. Literally, nôi» and ngoai » mean ‘inside’ and ‘outside’.5 These terms
emphasize the dominant status given to the paternal side in a patrilineal society
such as the Vietnamese.
Traditionally, a daughter is treated as an ‘outsider’ by her family, because
sooner or later she will belong to her husband’s family when she marries, as
» me» cha mua
reflected in the folk saying Con gái là con ngþǜi ta, con dâu mǘi thât
váe ‘Your daughter is someone else’s daughter, your daughter-in-law is really the
daughter that you bought’. This attitude also spreads to the generation of
cousins, as seen in the saying Con cô con câu » thì xa, con chú con bác thât
» là anh
em ‘Children of maternal siblings are not close relatives, but children of paternal
brothers are true brothers’. Consequently, Cháu bà nôi, » tôi
» bà ngoai» means the
child is a grandchild of the paternal grandmother, but taking care of the child
is the maternal grandmother’s obligation.
Table 1 shows that generally the finest distinctions are made in the parents’
generation, where the parents’ siblings are differentiated by both gender
(uncle/aunt) and lineage. The paternal uncles (but not the maternal uncles or
paternal aunts) are also differentiated by relative age, with father’s older brother
referred to as bác. The older brother’s wife is also referred to as bác. Unlike me»
‘mother/mom’, the term ba occurs only for address and not for reference. The
common noun for ‘father’ is cha. Depending on the dialect and social back-
ground of the speaker, e.g., social class, urban vs. rural, youngsters can use
»
many different expressions to address their parents, including bâo, cha, tháay, câu,
tía, bá, and bo» for the father and me, má, ma, » mü,» u, bu, and bam á for the
mother. Siblings of the speaker are differentiated by age, as older or younger
siblings, and the older siblings (but not the younger) are further differentiated
by gender. In a society that values social hierarchy, this shows the respect for
people in a higher position.
The gender of relatives is distinguished by trai ‘male’ and gái ‘female’, e.g.,
bác trai ‘father’s older brother’, bác gái ‘father’s older brother’s wife’; em trai
‘younger brother’, em gái ‘younger sister’; cháu trai ‘nephew/grandson’, cháu gái
‘niece/granddaughter’. However, great-grandfather is ông cố and great-grand-
mother is bà cố.
Married couples often employ kinship terms in addressing one another.
Typically, if the couple is young or middle-aged, the husband uses anh ‘older
brother’ for self-reference and em ‘younger sibling’ to address his wife, and the
wife uses em for self-reference and anh to address her husband. When the
couples grow old and have grandchildren, they use more balanced terms to
address each other. Wives use the expressions ông ‘grandfather’ or ông nó
Vietnamese 287
grandfather + he/she to refer to their husbands, and husbands refer to their

wives as bà ‘grandmother’ or bà nó grandmother + he/she. Both husbands and
wives use the pronoun tôi for self-reference. In urban settings, the couples often
use the kinship terms anh and em until they have grandchildren or until they
are very old. Other variations in address terms between couples are discussed in
detail in Section 5.4.
3.3 Pronouns
The choice of pronouns6 depends partly on formality, i.e., some occur in
formal contexts and others in colloquial contexts (spoken language, personal
letters, and written dialogue) where the speaker has a close or egalitarian
relationship with the addressees. The most common pronouns of Vietnamese
are listed in Table 2.
The first person singular tôi is used in formal or polite contexts to demon-
strate a considerable social distance between speaker and addressee; it originates
from a noun meaning ‘servant of the king’. In formal contexts, only tôi and ho»
(3pl) can occur. The terms tui and hÔan are colloquial Southern versions of tôi
and nó (3sg), respectively. The term ngài (2sg) is used in diplomatic and
religious contexts. A few forms, such as ta ‘I’ and ngþüi ‘you’ are archaic or old
words which occur in classical literature (e.g., fairy tales, historical novels, etc.).
The term ta can function as the first person plural inclusive, but usually in
written language. Other terms, such as chàng ‘he’ or nàng ‘she’, occur only in
Table 2.Vietnamese personal pronouns

Person Colloquial Formal/Polite Archaic/Literary
sg
1 tao tôi / tui (S) ta
2 mày ngài ngþüi
3 nó / hÔan (S) chàng (male)
nàng (female)
pl
1 ta / mình ta ta
2 bay bay bay
3 ho» ho» ho»
(S) = Southern usage

288 Hoa Pham
literary Vietnamese (poetry, novels, folk tales, lyrics, songs, etc.).

Among the third person pronouns, the gender distinction is made only in
the pair chàng/nàng. The rest can refer to both females and males. The third
person singular pronoun hÔan is used generically in spoken language, but in
written language, it usually refers to males. Ngài is generic as a religious term,
but in diplomatic contexts, the term ngài ‘you’ exclusively implies male. This
is because certain important positions or occupations were taken only by men:
politicians, physicians, professors, and lawyers. Today, if the addressee were
female, ngài ‘you’ would not be employed; instead, the kinship term bà ‘grand-
mother’ would be used, as in bà thu tþǘng ‘Mrs Prime Minister’, or bà Ðai » sþ́
‘Mrs Ambassador’. There is a generic pronominal term that originates from a
noun: mình ‘body’. In sentences such as the following mình is used as a
reflexive pronoun:
»
(1) Màoi hoc sinh phaiÑ mang sách cua mình.
each student must bring text book of self
‘Each student has to bring their own textbook.’
Mình can serve as a personal pronoun for both genders meaning ‘I, you (sing)’
and ‘we (inclusive)’ in colloquial language. The special use of this term will be
further discussed in Section 5.2.
Apart from the plural pronouns ho» (3pl) and bay (2pl), plurality can also
be indicated by the use of two plural markers, chúng and các. For instance, ngài
is pluralized by các, i.e., các ngài. The other singular pronouns can be pluralized
by chúng , e.g., chúng nó ‘they’, chúng ta ‘we (inclusive)’, chúng tôi ‘we (exclu-
sive)’, chúng mày ‘you (pl)’; see Emeneau (1951) and Thompson (1965) for
detailed discussions of personal pronouns.
The choice of a particular pronoun carries social significance beyond the
denotative meaning of the pronoun. A child can tell how upset his or her parent
is by the address term the parent uses. A husband can understand how his wife
feels about what he has said by her choice of words. An observer can presume
certain background information about the social class, education, individual
traits and origin of a person when they first meet by the use of address terms.
Names can also serve as address terms. They can be used to refer to the
speaker, the hearer or someone not present. Using proper names for self-
reference occurs more in female speech than in male speech (cf. Lþüng 1990:108).
In many cases, proper names are seen as neutral terms to avoid embarrassment
in situations where the social hierarchy is not clearly defined. This issue will be
discussed in Section 5.
Vietnamese 289
4. Male/female relations in idiomatic expressions
In Vietnamese, there are some idioms and folk sayings which denigrate women
in certain situations, but no similar ones exist for men in the same situations.
For instance, the abusive term gái già girl + old ‘old girl, spinster’ refers to a
woman who remains single after a certain age. If a married woman does not
have children, she may receive the comment:
(2) Gái 0ôc
» không con.
girl poisonous no offspring
‘Only bad women cannot have children.’
Referring to a married woman who already has children, the term na» dòng ‘a
woman having children’ carries a negative meaning, something like ‘a used-up
woman’, but there is no similar term for a married man. When a woman cheats
» cháong girl + mistaken + husband ‘an
on her husband, she is referred to as gái lôn
unfaithful woman’. There is no comparable expression for a man in the same
situation. These asymmetries clearly indicate a society where moral criteria are
much more strictly applied to women than to men. Similarly, if a woman is left by
her husband, something must be wrong with her, as indicated by the saying:
à phi chþ́ng no» thì tât
(3) Gái cháong ray, » kia.
girl husband leave not bad.lot this so bad.habit that
‘An abandoned wife must certainly have a bad character.’
Women are also judged by their physical attributes. For instance, a wide mouth
is a good thing in men but a bad one in women:
» rông
(4) Ðàǹ ông » miêng
» thì tài, 0àn bà
men wide mouth then talent women
» miêng
rông » 0iâec tai láng giáeng.
wide mouth deaf ear neighbor
‘A man with a wide mouth is stylish, a woman with a wide mouth is
noisy.’
If a woman’s hair is naturally curly, she must be a jealous person:

(5) Quăn quăn tóc trán là ngþǜi hay ghen.
curly hair forehead be person often jealous
‘Women with curly hair are often jealous people.’
290 Hoa Pham
If she is fat, she eats on the sly and beats her children:
»
(6) Ãn vung nhþ chǘp 0ánh con caÑ ngày.
eat on.the.sly like lightning beat offspring all day
‘(She) is very good at eating on the sly, beats her children all the time.’
These proverbs are taken from a compendium of Vietnamese folk sayings (Vũ
1994), but there are no such stereotypes about men’s obesity or their curly hair.
For a man to have a naive wife is regarded a disaster:
(7) Thþ́ nhâat vü» dai» trong nhà, thþ́ nhì trâu châm
» thþ́
firstly wife stupid in house secondly ox slow
ba rþa
» cùn.
thirdly bush-hook blunt
‘The worst thing is having a stupid wife, the second worst is having a
slow ox, and the third is having a blunt bush-hook.’
On the assumption that husbands are always smarter than their wives, the first
thing a man should do when he gets married is educate his wife:
» con
(8) Day tþ̀ thuüÑ còn thü, day » vü» day» thuüÑ
teach offspring since time still young teach wife teach time
bü vü mǘi váe.
forlorn/unprotected just come
‘Teach your child when he is still very young, teach your wife when
she first comes to live with you.’
When a young man is widowed, social sympathy is shown to single fathers by

the proverb gà trâong nuôi con ‘the rooster must raise the chicks’, but for single
mothers there is no comparable proverb, because this circumstance is seen as
normal and natural. The mothers are blamed, if their children do not behave:
(9) Con hþ tai» me.»
offspring bad due.to mother
‘A child has bad behavior if its mother does not do a good job.’
Although some of those views showing gender-related stereotypes have

changed, most of them still remain the same in the opinion of both men and
women, especially in the rural areas.
Vietnamese 291
5. Vietnamese terms of address, reference and self-reference
The Vietnamese system of address, reference and self-reference is a very

important area of discourse and it, too, shows many interesting asymmetries
related to gender. When people speak to one another in Vietnamese, they must
gauge several social factors if they are to avoid embarrassment or bad manners
in choosing terms of self-reference or reference to others. The following
examples, in which a mother addresses her daughter using three different
subsystems of address and self-reference, will give an impression of the com-
plexities of the address system.
In (10a) the mother employs the pronominal pair tao … mày ‘I … thou’.
In (10b) the mother uses a kinship term ‘mother’ to refer to herself and the
proper name Hiáen for her daughter. In (10c) the mother again refers to herself
in the third person as me» ‘mother’, but uses the kinship term con ‘offspring’ for
her daughter.
(10) a. Tao sẽ mua cho mày cái bút ay.â
I fut buy for thou cl pen that
‘I will buy that pen for you.’
b. Me» â
sẽ mua cho Hiáen cái bút ay.
mother fut buy for Hien cl pen that
c. Me» sẽ mua cho con â
cái bút ay.
mother fut buy for offspring cl pen that
All of these sentences mean ‘I will buy that pen for you’, although each uses
different forms for person reference.
The pair tao … mày usually occurs among close friends, children of the
same age, and, as in (10a) in the family by the elders towards younger people.
In (10b), the speaker uses the kinship term me» for herself and a proper name,
Hiáen, for her daughter. For parents, the use of proper names in addressing
their children occurs mainly in urban or educated families. In a conversation,
a change in reference terms from me» … con ‘mother … offspring’ in (10c) to
tao … mày in (10a) is usually a warning sign for the child that their parent is
not pleased.
5.1 Sociopolitical background and characteristics of address terms

Up to the early decades of the 20th century, social relations in Vietnam were
relatively static. People generally had a clear notion of their own status and of
292 Hoa Pham
the relative status of the people with whom they came into contact. The use of
address terms, though complex, was relatively well-defined.
As the 20th century progressed, a number of social changes created a more
fluid social system. Urbanization led to closer contacts among castes, classes and
social groups, and occupational mobility resulted in greater flexibility in social
status. Women, whose roles were formerly well-defined as largely domestic and
reproductive, began participating in the more public spheres of education and
later business, commerce, and politics. To some extent, these changes were a
reflection of the influential French colonial system, which provided a European
perspective for at least the upper echelon of Vietnamese society. Vietnam
encountered additional social forces for change: Many years of civil war brought
the country into close contact with occidental cultures, not only French, but
also American. These contacts conspired with the other social changes to
“democratize” Vietnamese society, which in turn affected the Vietnamese
system of person reference.
Women are more restricted in the terms they use, and are thus more
sensitive to their social effect. When choices of address terms change in any of
the subsystems described above, those changes are more likely found in the
speech of women than men.
One of the essential characteristics of the system is that the terms are highly
context-sensitive. When speakers choose terms of address, self-reference and
reference in Vietnamese, they must be aware of the social nuances of the
situation. Consequently, there is a tendency to avoid certain terms in specific
contexts. For example, in the first half of this century, in rural areas, a young
wife might avoid using any specific terms to refer to herself or to her husband.
The conversation below, which is a well-known witticism in Vietnam, illustrates
this avoidance. In (11), the young wife addresses her husband using ai, an
indefinite pronoun meaning ‘someone, everyone, anyone’ in a statement or
‘who’ in a question. In the husband’s reply, he teases his wife by using the literal
meaning of the pronoun. In the last utterance, the first two ais refer to the wife
and husband, respectively; the second two ais have the literal meaning of an
interrogative pronoun ‘who’, which is perfectly understood.
(11) Wife: Ai üi, váe ăn cüm!
someone voc return eat rice
‘Hey, someone, go home for lunch!’
Husband: Ai goi » ai 0ay?â
who call who prt
‘Who is calling whom?’
Vietnamese 293
Wife: Ai » ai
goi chþ́ còn ai goi
» ai nþ̃a?
someone call someone prt still who call who then
‘I (someone) am calling you (someone), who else (who) is
calling you (who)?’
The use of some common nouns as terms of address or self-reference with

special meanings is quite common. For instance, in (12), a young woman refers
to herself as ngþǜi ta ‘people’ when speaking to her boyfriend:
(12) Hôm qua không tǘ¯i làm ngþǜi ta chǜ mãi.
day pass not come make people wait ever
‘I (people) waited (for you) for a long time yesterday but (you) did
not come.’
Here, the girl is blaming her boyfriend for not coming. She uses ngþǜi ta for
self-reference and no term at all for the addressee. The unmarked use in this
case would be the kinship terms anh ‘elder brother’ for the addressee and em
‘younger sibling’ for self-reference (Hôm qua anh không tǘi làm em chǜ mãi).
These kinship terms would show normal intimacy, but that is not what the girl
wants. She wants to convey the message that she is upset, so she chooses a term
that literally refers to herself as a third party ngþǜi ta and omits any explicit
reference to her boyfriend.
Depending on the context, ngþǜi ta can be used to refer to the addressee or
a third person, as in (13). Here, a girl speaks to her boyfriend about another girl
whom she considers her rival. She uses tôi for self-reference to show aloofness
and jealousy, the first ngþǜi ta for that girl (her rival), and the second ngþǜi ta
for the addressee:
(13) Tôi 0âu có 0ep » nhþ ngþǜi ta 0ße 0þüc
» ngþǜi ta
I not have beauty like people in.order.to get people
ngó tǘi!
look.at prt
‘I know that (I am) not as beautiful as her (people) to get your
(people’s) attention!’
Although the speaker in (12) and (13) can be female or male, women are more
likely to speak this way than men. The speakers indicate a certain deliberate
distance in a close relationship by referring to themselves or to the addressee
with a common noun otherwise used for a third party, ngþǜi ta ‘people/they’,
in intimate contexts.
294 Hoa Pham
5.2 Kinship terms and common nouns

With a limited set of personal pronouns speakers of Vietnamese have to use
kinship terms and common nouns as address terms. Here, too, the address
terms can easily cause communicative and social problems if they are used
inappropriately.
Kinship terms can be used with people who are not the speaker’s relatives.
In informal contexts, speakers must choose a kinship term appropriate to their
age and relationship with the addressee. For example, in (14), the speaker
addresses her older neighbor using the term cháu ‘grandchild/nephew/niece’ to
refer to herself and dì ‘mother’s sister’ for her neighbor:
(14) Cháu mǘi mua cho dì Tþ cái lþüc » ngô» lÔam.
niece just buy give aunt Tu cl comb cute very
‘I (niece) have just bought you (mother’s sister Tu) a very cute comb.’
The addressee can be referred to as dì, if she is older than the speaker, but
younger than her grandparents.
Among the kinship terms, the two that occur in the most formal contexts
are ông ‘grandfather’ and bà ‘grandmother’. In addressing people a generation
or more older, speakers use the terms ông with the general meaning ‘Mr, Sir’ or
bà with the meaning ‘Mrs, Madam’. In addressing younger people or in less
formal circumstances, speakers can make use of the expressions anh ‘elder
brother’ or ch»i ‘elder sister’.
Kinship terms allow the speaker to convey various feelings or attitudes to
addressees who are not relatives. Maternal terms are usually regarded as more
intimate than paternal terms. For example, if dì ‘mother’s sister’ in (14) were
replaced by cô ‘father’s sister’, the speaker would display a greater distance from
the addressee. This is because for second and third generations, maternal
relatives usually are associated with care, gentleness and unconditional love,
while paternal relatives are seen as strict and powerful rulers.
Certain common nouns can serve as address terms. The speaker can refer
to the addressee by a common noun, as in (15a) or by a noun indicating the
occupation or status of the addressee, as in (15b). The addressee in these
sentences might be the same person, but (15a) shows greater intimacy between
the speaker and the addressee:
(15) a. Tôi thayâ ban
» có vÑe thiâeu nguÑ 0ó.
I see friend have appearance lack sleep prt
‘I see that you (friend) seem not to sleep enough.’
Vietnamese 295
b. Tôi thayâ bác sı̃ có vÑe thiâeu ngþ 0ó.

I see doctor have appearance lack sleep prt
‘I see that you (doctor) seem not to sleep enough.’
In (15a) the speaker uses ban» ‘friend’ to refer to the addressee, thereby showing
the friendship that exists between the two of them. The speaker may be older or
the same age, but not younger than the addressee. In (15b), the speaker uses the
occupational term bác sı̃ ‘doctor’ to refer to the addressee, displaying greater
distance in their relationship. The speaker might be older or younger than the
addressee, and is not necessarily a patient.
The term mình ‘body’ can occur in many different relationships, such as
between friends and married couples. It is used as an address term among
married couples and for self-reference in other relationships. The term occurs
only in solidarity contexts or fairly non-formal contexts, such as between two
employees who have worked with each other for a long time.
Between married couples, mình is used to address the spouse, by either the
husband or wife. If the speaker is the husband, he uses anh ‘elder brother’ for
self-reference. If the speaker is a wife, she uses em ‘younger sibling’ for self-
reference. For example, in (16), the husband speaks to his wife:
(16) Anh 0ã nói bà Phán sáng mai
elder.brother past tell grandmother Phan morning tomorrow
» mình.
găp
meet body
‘I (elder brother) told Mrs Phan to see you (body) tomorrow morning.’
There are no age constraints for the term mình. This use is found mostly in
urban settings. Among female and male friends, mình ‘body’ can function as
self-reference when speaking to anyone who is younger or the same age. If the
speaker were younger than the addressee and used this term for self-reference,
it would be considered inappropriate and impolite. If the speaker were much
older than the addressee, the speaker’s use of mình would show a greater degree
of intimacy. For mình to be used when the ages differ, the speaker and the
addressee usually have to share the same work environment, such as employees
who do the same job in the same company, or professors and students. In the
relationship between teachers and students, a young or “non-conservative”
teacher might use mình for self-reference when speaking to students outside the
classroom in senior high school, colleges or universities. In this context, the
teacher wants to show solidarity with the students rather than authority. This
use is not allowed in formal contexts.
296 Hoa Pham
Between females or when speaking to a person of the opposite gender, a

speaker who uses mình for self-reference usually employs a proper name to
address the interlocutor. This usage is common among young and middle-aged
people in urban settings.
Between males, mình can be used for self-reference when speaking to a close
friend, or to a person the speaker first meets in non-formal contexts who is
younger or the same age. However, if the conversation is between a male and a
female, they need to be close friends in order to use the term mình for self-
reference. If they are not close friends, the man would be regarded as impolite
regardless of his age, and the woman might be regarded as flirtatious if she was
about the same age. However, this usage is acceptable if the woman is much
older than the man.
In Southern dialects, except among married couples, speakers use the term
mình only for self-reference, but in North Vietnam, mình can refer to either
the speaker or the addressee. This leads to different possible meanings for
sentences like (17):
» ông Tam sáng
(17) Mình sẽ găp mai.
body will meet Mr Tam morning tomorrow
A Southerner would understand that it is the speaker who will meet Mr Tam
tomorrow morning; but the sentence is ambiguous for a Northerner, since it
can also mean ‘You will meet Mr Tam tomorrow morning’.
Among friends, tǘ ‘servant’ can occur as the first person to show a close
relationship. A friend speaks to another friend in (18):
»
(18) Câu 0þ̀ng nói nþ̃a, tǘ chán lÔam ráoi!
mother’s.brother not say more servant fed.up much already
‘You should not speak any more, I (servant) am already fed up!’
Sentence (18) can be used by female as well as male speakers. The kinship term
» though it refers explicitly to a male, i.e. ‘mother’s brother’, can be em-
câu,
ployed in Northern dialects to address a female friend.
In Southern dialects, the term corresponding to Northern câu » and 0Óang aâ y
side + that ‘you’ is báo ‘large basket’. It means something like English ‘buddy’.
Generally this term is used between people of the same age and it is very
common, especially in female speech, perhaps because women show intimacy
more easily than men and are more open towards friends.
Another popular term among Saigonese youngsters is nhoÑ ‘little’. It can be
used to refer to someone who is younger or the same age. In (19), the speaker
Vietnamese 297
is talking to her friend about another girl or woman who might not be their
friend. The speaker uses tui for self-reference, nhoÑ ‘little’ for the third person,
and báo ‘large basket’ for the addressee:
â nhoÑ 0ó coi bô»
(19) Tui thay siêng hün
I see little that see appearance hard.working than
báo nghen.
large.basket prt
‘I think that girl (little) seems to be more hard-working than you
(large basket).’
The term nhoÑ ‘little’ can also refer to the interlocutor. For example, a female
speaks to her female friend in (20):
(20) Mai nhoÑ nhǘ chǜ ta 0i hoc
» vǘi nghe.
tomorrow little remember wait.for I go study with prt
‘Tomorrow you (little) remember to wait for me to go to school.’
The pair ta … nhoÑ ‘I … little’ as used in (20) is very common between female
students. Male students do not use the term nhoÑ to address male friends, but
only to address close female friends or mostly girlfriends. In speaking to a lover,
a male speaker can select the kinship term anh ‘elder brother’ for self-reference.
In this case, nhoÑ means something like ‘sweetheart’ or ‘honey’ in English. The
term nhoÑ in this situation can be replaced with bé ‘small’ with the same intima-
cy. The term bé resembles the meaning of English ‘baby’. Young children can
also be addressed as bé.
In the next section, I will show that terms of address, self-reference and
reference have changed in contemporary Vietnamese society, especially the
terms women use in all types of relationships. Although the sociocultural
organization of Vietnam differs from Western countries, the sociolinguistic
behavior of females and males is susceptible to the same pressures for variation
and change.
5.3 Terms of address and self-reference in various female–male

encounters
Beginning in the first decades of the 20th century, partly under the influence
of new values brought by the French, there was a social revolution against
traditional values and rules. Its effects are especially noticeable in asymmetric
power relationships such as those between masters and servants, teachers and
298 Hoa Pham
students, and husbands and wives. The differences also pervade symmetrical
relationships, such as those between neighbors and work mates. People have
been led to re-evaluate their traditional roles. Equality and inclusiveness, particu-
larly for women, have been on the increase slowly but surely for several decades.
The beginnings of this revolution were vividly reflected in popular literature
during the period of 1930–1945, in the first Vietnamese novels. The novels
depicted the new social mores, sometimes through characters who were socially
advanced but more often through characters who were intended to represent
the contemporary urban class. Above all, the development of the Vietnamese
novel introduced realism into literature. The interaction of the characters
preserves their attitudes and, more important, the language they used to express
those attitudes. Under the new view, some terms came to be considered old-
fashioned, and are now nearly obsolete. New ways of address and self-reference,
which are more relaxed and open, came into being. I will briefly discuss some
ways in which these traditional subsystems have changed to accommodate
social changes, especially in regard to gender roles.
5.3.1 Masters and housekeepers

In traditional society, relationships such as those between housekeepers and
masters, students and teachers, or wives and husbands, were strictly hierarchi-
cal. The hierarchy was codified in the use of terms of address and self-reference.
Housekeepers usually employed con ‘offspring’ to refer to themselves when
speaking to their masters, and used terms such as ông ‘grandfather’ or bà
‘grandmother’, or something similar, to address the master or mistress, regard-
»
less of the age of the individuals involved. They used cô ‘father’s sister’ or câu
‘mother’s brother’ to address the offspring of the master. Masters and their
offspring, however, usually utilized the pair tao … mày ‘I … thou’ when
speaking to their housekeepers.
In Vietnamese society before 1954, housekeeping was a common occupa-
tion. Middle class families might employ several people to do housework:
women to cook, babysit, or clean the house, and men to garden and drive.
Lower middle class families usually employed a woman as cook, nanny and
cleaner in the house. The housekeeper was generally a woman or a girl, and
housekeeping has always been and still is regarded as women’s work.
A short story from the 1930s, Cái Ve by Khái Hþng, dramatizes the chang-
ing relations between servants and masters. An 18 year-old girl cooks for some
tenants who are about the same age or a few years older. Some of them are
workers, one is a young teacher. In (21a), the girl speaks to the teacher, whom
Vietnamese 299
she likes very much. She uses con ‘offspring’ to refer to herself and tháay ‘teacher’
to address the teacher:
(21) a. Girl: Con 0ße 0ây, thay
á chǜ môt
» lát
offspring put here teacher wait a moment
â
chè ngam 0ã ráoi hãy uâong.
tea absorbed already then will drink
‘I (offspring) put (tea) here, you (teacher) should wait
for a minute until it is ready.’
b. Teacher: Ñ cô
Tôi bao 0þ̀ng xþng con
I tell father’s.sister not address.oneself offspring
vǘi tôi, sao cô không nghe?
with me why father’s.sister not listen
‘I have told you (father’s sister) not to use ‘offspring’
when speaking to me, why don’t you (father’s sister)
listen?’ (Khái Hþng 1936: 23)
In (21b), the teacher uses the pronoun tôi to refer to himself and cô ‘father’s
sister’ to refer to the girl.
In (21), the teacher, as an educated person, is attempting to bring the spirit
of the social revolution to uneducated people like the girl. He expects the girl to
use a more appropriate term to refer to herself according to her age, but not the
term con which is regarded as self-deprecating. He uses the term cô ‘father’s
sister’ to address the girl. The literal meaning of this kinship term has been lost
here. It is used to address young girls or women of the same class as the teacher.
As the scene progresses, it becomes clear that this expression upsets the girl
because she is not used to it. The novelist tells us that the girl thinks ‘that term
should be used only for elegant, beautiful people in noble and rich families. She
first thought it was a bitter mockery’ (Khái Hþng 1936: 13). However, later in
their relationship, the girl selects ông ‘grandfather’ to refer to the teacher and tôi
‘I’ to refer to herself.
After North Vietnam became the Democratic Republic of Vietnam in 1954,
housekeeping positions were abolished. In South Vietnam, housekeeping was
still considered a suitable job for young girls and women until 1975 when
Vietnam was reunited. From 1954 to 1975, servants used terms appropriate to
their age to refer to themselves when speaking to their masters. The female
servants used the neutral term tôi ‘I’ if they were older or the same age as the
masters, em ‘younger sibling’ if they were somewhat younger, and con ‘off-
spring’ if they were about the same age as the masters’ offspring. Under present
300 Hoa Pham
economic reforms, housekeeping has again become a desirable job in Vietnam.

Not only do housekeepers use terms appropriate to their age to refer to them-
selves, but their employers no longer use the pair tao … mày ‘I … thou’ when
speaking to their housekeepers as they did before, unless the context is appro-
priate, i.e. when the speaker is older but wants to show familiarity between the
two of them. Masters now, depending on their age, can use tôi ‘I’ or con
‘offspring’ to refer to themselves, and ch»i ‘elder sister’, dì ‘mother’s sister’ or
something similar to refer to the servant. The servant can now use the term tôi
or con to refer to herself when speaking to the masters. In addressing young
» ‘mother’s brother’.
masters, the housekeeper can use cô ‘father’s sister’ or câu
5.3.2 Teachers and students

When a teacher is talking to younger students in formal contexts, a male can
use tháay ‘master’ and a female can employ cô ‘father’s sister’ for self-reference.
This usage reveals another asymmetry, because women cannot be referred to as
tháay even though the word literally means, simply, ‘teacher’. When the term is
applied to women, it is for fortune-tellers instead of teachers, and in this
instance it is always qualified as bà tháay, that is, the gender must be specified as
female even though the word is not marked as male.
In many places, when speaking to teachers who are as old as their parents,
students use con ‘offspring’ for self-reference. The term con ‘offspring’ was very
common before 1975 in South Vietnam, and especially before 1954 in both
North and South Vietnam. Today, the most common term used by students for
self-reference is em ‘younger sibling’. Using tôi ‘I’ when speaking to teachers is
extremely rare, even among adult students, since it is considered too egalitarian
in this relationship.
Currently, students can also refer to themselves by using proper names
when speaking to their teachers. This use was not possible in the past when
students used only con ‘offspring’ to refer to themselves. Using proper names
for self-reference occurs mainly among young students and female students, but
not among older male students in either North or South Vietnam, because this
is considered “feminine”.
Under current social conditions, address and self-reference sometimes pose
a problem for adult male students. Before 1945, it was very rare for Vietnamese
women to receive higher education or any formal education at all. After 1954,
when Vietnam became independent, women slowly achieved almost the same
rights to education as men. Many young women now become post-secondary
school teachers. At the same time, many people, especially men, have a chance
Vietnamese 301
to go back to school in their middle age.

Adult male students can easily use the typical term em ‘younger sibling’ to
refer to themselves when speaking to a male teacher, whether the teacher is
older or younger. However, when the teacher is female and younger than the
male student, it is extremely difficult for the student to choose a term for self-
reference. The traditional term for addressing female teachers, cô ‘father’s sister’
is awkward. If the student is much older than the female teacher, the student
can use tôi ‘I’ for self-reference. If the student is not very much older, the
pronoun tôi does not sound polite enough, because the pair cô … tôi is usually
employed by a male speaking to a younger girl when the man is socially
superior. In this case, if he were to use the term em ‘younger sibling’, it would
diminish his masculinity. Men in this position usually resort to either using
their proper names or avoiding any terms at all for self-reference. Sometimes
adult male students use the term cô giáo father’s sister + teach = ‘female teacher’
to refer to the female teacher and the term hoc» trò study + person = ‘student’ to
refer to themselves. This term shows intimacy in a pleasant and rather humor-
ous way, because the term hoc » trò is normally used only to refer to elementary
students, not to adult or university students, and the term cô giáo ‘female
teacher’ is an occupational term that is not normally used as an address term.
Female adult students have no problem using the term em when speaking
to younger teachers, whether female or male. Avoidance usually occurs exclu-
sively in the speech of male adult students in this case.
Teachers can also address their students by their proper names. Using
proper names shows greater intimacy in this relationship. For instance, in (22),
a teacher speaks to his student, Nam, by name:
(22) Ngày mai tháay sẽ 0þa Nam ra bâen 0ò.
day tomorrow teacher fut lead Nam to landing ferry
‘Tomorrow I (teacher) will go with you (Nam) to the ferry-landing.’
The student in (22) might be a male or female, child or adult, but must be
younger than the teacher. In referring to himself by his occupation, thay á
‘teacher’, the teacher is using the most neutral of all terms for self-reference.
Female teachers use proper names more than male teachers; such usage is
regarded as more gentle and intimate in solidarity contexts.
5.3.3 Neighbors and work mates

In informal contexts, saying em ‘younger sibling’ for self-reference when
speaking to an acquaintance or neighbor is quite common among women in
302 Hoa Pham
both North and South Vietnam, particularly when speaking to men. Addressees
usually receive bác ‘father’s elder brother’ in rural settings, and anh ‘elder
brother’ or ch»i ‘elder sister’ in urban settings. Men, by contrast, rarely use the
term em ‘younger sibling’ for self-reference when speaking to women, unless
they are considerably younger than the women.
However, for many young women in urban South Vietnam, using the
expression em for self-reference when speaking to a male with whom they are
not very familiar might be considered to be humbling or indicate low self-
esteem. In this situation, the term tôi, on the other hand, might be regarded as
pompous or too formal. Again, proper names are usually preferred to avoid
such difficulties. The term em is reserved for self-reference when people are
talking to someone they really regard as an ‘older brother’: older men in close
relationships, such as their close friends or their older brothers’ friends.
Women sometimes use both terms, em ‘younger sibling’ and tôi, in differ-
ent styles. In (23a), a woman in her late 30s speaks to an older man in his 50s in
the living room, after they have been introduced to each other by another man,
a mutual friend, who is in his late 40s. The woman is a writer, and the man she
is meeting is a magazine editor, and a potential publisher of her work. The
woman uses em for self-reference, since the man is quite a bit older, and she
wants to show intimacy in a relaxed context. She employs anh ‘elder brother’ to
refer to both men:
(23) a. Em nghe anh â nói váe
Tan anh
younger.sibling hear elder.brother Tan talk about elder.brother
hoài »
mà bþ̃a nay mǘi găp.
all.the.time that today just meet
‘I (younger sibling) hear (elder brother) Tan talk about you (elder
brother) all the time, so today we meet.’
In (23b), the man replies, using ch»i ‘elder sister’ to address the young woman
and tôi for self-reference:
(23) b. Cám ün ch»i. â mong
Tôi cũng rat 0þüc»
thank elder.sister I also very look.forward get
» ch»i.
găp
meet elder.sister
‘Thank you (elder sister). I also looked forward to seeing you (elder
sister) very much.’
Later, when they discuss the possibilities of his publishing her article in his
magazine, the woman changes the form of self-reference:
Vietnamese 303
(24) Tôi 0ã nhân » 0þüc

» sâo báo mǘi nhatâ
I past receive get number newspaper new most
anh gþi
Ñ tuaná ráoi.
elder.brother send week last
‘I have received the latest issue you sent last week.’
The woman changes from em for self-reference in a relaxed conversation as in

(23a) to a formal term, tôi, when they talk about work in (24). This use of the
term tôi is perfectly acceptable, because age does not play a crucial role at work.
Between women, however, using tôi for self-reference at work can some-
times create difficulties. This term is used in formal situations to maintain social
distance. When a woman speaks to a younger woman or a woman occupying a
lower social position, and wants to eliminate the distance, she is faced with a
difficult situation: The speaker wants to avoid using the term tôi, but at the
same time wishes to maintain a certain politeness and formality. Because of
situations like these, a new pair of self-reference and address terms has been
coined: mình … ch»i ‘body … elder sister’. In (25), a female family doctor speaks
to her younger patient in her office in Toronto, using mình ‘body’ to refer to
herself and ch»i ‘elder sister’ to address the patient:
(25) Mình cho ch»i thþ́ thuâoc khác coi có hün không.
body give elder.sister kind medicine different see have better prt
‘I (body) give you (elder sister) a different kind of medicine to see whether
that works better.’
Politeness still remains a feature in the address term ch»i ‘elder sister’ the doctor
uses to address the patient. The use of mình ‘body’ in this way is very recent and
still not common, but it now seems to be accepted. It can be heard from time to
time in relationships not only between a family physician and her patient but
also, e.g., a pharmacist and her new customer, or two educated women who
have common interests.
In the North, when communism became well established, a new expression,
0áong chí ‘comrade’ (literally, one with the same purpose or will) became very
common for addressing women and men of all ages and positions in order to
show equal status. In many cases, this term even occurred in personal conversa-
tions between two friends, though never between family members.
The same ideological pressures led to traditional terms being regarded as
“remnants of feudalism” or “bourgeois”. They were eliminated from daily life
in public communication and school. Consequently, the diplomatic and very
formal term ngài ‘you’ to address a male person was completely eradicated,
304 Hoa Pham
since it was considered to be an exclusive term for people of the dominant class.
However, when the current Vietnamese socialist government had to carry out
economic reforms and open its economy to the Western world in the late 1980s
and especially in the 1990s, the term ngài ‘you’ returned as an address term in
diplomatic contexts.
In the last two decades, the term 0áong chí ‘comrade’ has gradually disap-
peared. It became a relic of the war years. Instead, kin terms such as ông
‘grandfather’, bà ‘grandmother’, anh ‘older brother’ or ch»i ‘older sister’ are
now used in public communication situations such as television, radio and
newspapers.
5.4 Women and men as couples

Among married couples, more intimate address terms have come into common
use. In traditional Vietnamese society, and still in many parts of present-day
Vietnam, husbands were masters in their families. They made all the important
decisions and had great power over their wives. Women were supposed to obey
their husbands in every situation. When guests came to visit or to dine, wives
worked in the kitchen and were not allowed to eat at the same table with their
husbands and their guests.
Before 1945, husbands could legally have more than one wife. The first
wife could buy a concubine for her husband in order to have a servant in the
house or to bear her husband’s children if she herself as the first wife did not
have any sons.
The asymmetry in the relationship between husband and wife was reflected
in the terms of address, reference and self-reference they used. The husband
generally could use the pair tao … mày ‘I … thou’ when speaking to his wife in
normal situations, but a wife was never allowed to select that pair when
speaking to her husband, since this was considered too familiar. In (26), taken
from the 1939 novel Làm Lẽ (‘Being a concubine’), a husband speaks to his
concubine about his first wife:
(26) Tao chÑı có thâe, vì tiáen lþüng cô mày
I only have that since money salary father’s.sister thou
Ñ
giþ̃ ca.
keep all
‘I only have that much, because my first wife (your aunt) keeps all mon-
» 1939: 144)
ey.’ (Manh
Vietnamese 305
Here the husband uses the term tao ‘I’ to refer to himself, mày ‘thou’ to refer to
the concubine, and cô ‘father’s sister’ to refer to his first wife. The kinship term
cô is also the term the concubine uses to address the first wife.
If the concubine had children, her children would be obliged to call the
first wife me» ‘mother’ and they would call their biological mother ch»i ‘elder
sister’, because the concubine’s role was to give birth for the husband and his
first wife. In many cases, the children of the concubine use the term bé ‘little’
to refer to their biological mother, which is also the term the first wife employs
to refer to the concubine.
Since 1954, there has been a law forbidding men to have more than one
wife. However, mainly in the South, some men still have concubines. Today,
children of the concubine use the term me» ‘mother’, not ch»i ‘elder sister’, to
refer to their biological mother as well as to the first wife.
Another expression that reflects the asymmetry in the relationship between
husbands and wives is the third person singular pronoun nó ‘he/she’. A man can
use this term to refer to his wife in normal situations, but again it is forbidden
for a wife to make use of it in referring to her husband. Even among many
young and educated people, this asymmetry still holds. The husband’s use of nó
‘she’ in public might be a way to show off the power of the husband over his
wife in front of a third party, or at least to show he is the decision-maker in his
family, whether or not this is true. However, today men use the term nó less
often. They usually employ polite terms to refer to their wives in the third
person, such as nhà tôi ‘my house’, bà ay â ‘that grandmother’. Sometimes they
combine the two; for instance in (27), the man uses nhà tôi ‘my house’ and nó
‘he/she’ to refer to his wife:
Ñ nhà tôi, nhþng nó
(27) Tôi hoi Ñ anh
bao không muâon
I ask house I but he/she say elder.brother not want
bán chiâec xe cũ.
sell cl vehicle old
‘I asked my wife (my house), but she said you (elder brother) did not
want to sell the old vehicle.’
Among working class people, this use of terms is quite common and is consid-
ered polite. However, a wife cannot use the same terms in that way to refer to
her husband, although both expressions, nhà tôi ‘my house’ and nó ‘he/she’, can
be employed to refer to a husband and a man, respectively. Instead of the
pronoun nó, she must say nhà tôi or ông ay â ‘that grandfather’, or something
similar. In this respect, the asymmetry between husbands and wives is preserved
in the working class.
306 Hoa Pham
In Vietnamese, kinship terms may be used as terms of address. The degree

of respect or intimacy in an address term depends on the kinship term used. A
man can be referred to by a kinship term followed by his proper name. For
example, ông Tam grandfather + Tam = ‘Mr Tam’. In (28), a speaker talks
about a man he knows:
(28) Thþ́ bÑay ráoi tôi găp
» ông Tam üÑ câulacb
» ô.
»
Saturday last I meet grandfather Tam in club
‘Last Saturday I met Mr (grandfather) Tam at the club.’
» Tam mother’s brother + Tam, ‘uncle
ông Tam in (28) can be replaced by câu
Tam’, which shows a closer relationship between the speaker and the person
talked about.
The kinship term in (28) may be replaced by special gender-specific
designations, thÓang for male and con for female. These terms can indicate
intimacy between the speaker and the addressees or persons talked about, if the
speaker is about the same age or older. For instance, in (29) a friend talks
about a couple:
(29) ThÓang Chín vǘi con » chÔac
Hanh là tháng
male Chin with female Hanh probably be month
sau cþǘi.
after marry
‘Chin and Hanh probably will get married next month.’
The speaker uses thÓang and con to refer to the couple in (29), thereby display-
ing intimacy. However, if the speaker is younger than the couple, these
designations indicate that the speaker does not like the couple. The two
designations thÓang and con can also be chosen by parents to refer to their
children to show intimacy.
The special classifiers (cl) thÓang and con can be used to refer to people
having low social status, such as con sen cl + servant ‘a female servant’, con 0iâem
cl + prostitute ‘a female prostitute’, or disreputable people such as thÓang ăncÔap
cl + steal ‘a male thief ’. Sometimes speakers indicate their feelings by their use
of these classifiers. For example, in (30), a woman speaks to her friend about a
situation in her company:
(30) Tháng trþǘc thÓang 0âoc công tþ» ý nâng lþüng cho
month previous male manager self will raise salary for
con Thu.
female Thu
‘Last month the manager himself increased the salary for Thu.’
Vietnamese 307
The speaker uses thÓang for the manager and con for the woman named Thu, and
by doing so, demonstrates her disapproval of the manager and jealousy of the
woman. In a normal situation, the speaker would use the kinship terms ông
‘grandfather’ and bà ‘grandmother’, or something similar, to refer to the
manager and the woman, respectively.
The special classifiers thÓang and con occur only in informal contexts. In
1975, Southerners were uncomfortable at first when officials of the new regime
referred to the officials or leaders of the former government with thÓang in
public meetings and the media, as in thÓang nguy » cl + rebellious/illegal =
‘official of the former government’, thÓang Thiêu» cl + name of the president of
South Vietnamese government. This use of the term quickly disappeared from
public communication. However, the term thÓang M«y cl + American ‘(male)
Americans’ remained in use much longer.
Among married couples or lovers, men can use the con to refer to their
wives or girlfriends. In (31), a man speaks to his neighbor who is sick:
(31) Ðße cháu bao Ñ con » sang nau
Hanh â cüm
let nephew tell female Hanh come cook rice
cho bác.
for father’s.elder.brother
‘Let me (nephew) tell Hanh to come over to cook for you (father’s elder
brother).’
The speaker employs the term con followed by a proper name, Hanh, » to refer to
his wife. This use is regarded as normal, even among young and educated men
in urban areas. However, if the speaker in (31) were a woman, she could not
speak about her husband by using the term thÓang followed by his name – even
if she were about the same age as or older than him – because this would seem
to overturn the family hierarchy. She would be regarded as impolite, because
the designation implies lower status.
Today, using con to refer to a wife or girlfriend is less common. Instead,
husbands often refer to their wives by the term vü» tôi wife + I ‘my wife’, or they
use frozen metaphors such as nhà tôi house + I ‘my wife’ (lit. ‘my house’) or bà
xã tôi grandmother + village + I ‘my wife’ (lit.‘my village grandmother’). The
metaphor ‘my house’ can also be used by a wife to refer to her husband. In the
term bà xã and the equivalent term for ‘husband’, ông xã grandfather + village,
the literal meanings are completely lost. These are relatively new expressions,
and they have become very common in both urban and rural areas among
young and middle-aged couples. They are used only in intimate contexts. For
instance, the speaker in (32) tells his friend:
308 Hoa Pham
(32) Tôi nghı̃ ông nên Ñ lai

hoi » bà xã
I think grandfather should ask again grandmother village
chuyên» này.
matter this
‘I think you (grandfather) should check with your wife (grandmother
village) again about this.’
The speaker uses ông ‘grandfather’ to refer to his friend and bà xã grandmother
+ village ‘wife’ to refer to the friend’s wife. The speaker in (32) may be young or
middle-aged.
In Vietnam, girls bear their father’s last names. Wives do not legally adopt
their husbands’ names after they get married, but they are usually referred to by
their husbands’ first names preceded by a kinship term: bà Vinh grandmother
+ Vinh ‘Mrs Vinh’ (the wife of Vinh), dì Nhiêm » mother’s sister + Nhiem ‘Mrs
Nhiem’ (the wife of Nhiem). Today, when many women work outside the
family, have a career and their own social status, wives are more often referred
to by their maiden names (i.e., their former names).
In urban settings, husbands can use proper names without a kinship term
or other qualifier to address their wives and also to refer to them when speaking
to friends and acquaintances. In rural areas, where women either do housework
or work in the field, husbands rarely use maiden names to refer to their wives,
and this usage never occurs with middle-aged and old couples.
Among young and middle-aged couples, when the kinship terms anh and
em are used for addressing each other, they do not pose any problem if hus-
bands are older than their wives. When husbands are younger than their wives,
we now find women changing the way in which they refer to their husbands
and themselves. In years past, when wives were normally homemakers and their
husbands were breadwinners, wives considered themselves to be under their
husbands’ protection. Hence, using the term em ‘younger sibling’ to refer to
herself was natural for a wife, whether or not she was the younger in the
marriage. The wife automatically assumed a lower status in the family: That was
a woman’s place. In urban settings now, if husbands are younger than their
wives, wives – particularly educated ones – consider the term em for self-
reference to be either humorous or embarrassing. In this case, proper names
come to the rescue: Wives refer to their husbands and to themselves by proper
names, or they use anh ‘elder brother’ to address their husbands and their own
proper names to refer to themselves.
In referring to their girlfriends, men usually use only proper names without
the special designation con. With young couples today, using proper names has
Vietnamese 309
become very common so that among urban people, women sometimes use
proper names to address older boyfriends and the term em ‘younger sibling’ for
self-reference:
(33) Chiáeu qua em tính tǘi Hþng mà trǜi mþa quá.
afternoon pass younger.sibling intend come Hung but sky rain much
‘Yesterday afternoon I (younger sibling) wanted to come to see you
(Hung), but it rained so heavily.’
The girl in (33) uses em ‘younger sibling’ to refer to herself, and the man’s
name, Hung, to refer to her boyfriend. This use of proper names would have
been regarded as inappropriate a generation ago, because when addressing an
older person, the speaker was expected to use an appropriate kinship term
followed by the proper name, such as anh Hþng ‘elder brother Hung’, chú
Hþng ‘father’s younger brother Hung’. Proper names without a preceding
kinship term were used only to address younger persons. In (33), by omitting
a kinship term before her boyfriend’s name, the speaker erases the hierarchy
implied in age differences and makes the relationship as between best friends.
However, she shows her respect for her boyfriend in her use of em ‘younger
sibling’ for self-reference. Her use of his proper name is innovative, and a
positive indicator of equality.
Intimate address terms such as cþng ‘sweetheart’ or bé ‘little, baby’ have also
become common among young couples. Although these terms are mainly used
by men to address their wives or girlfriends, young women sometimes use cþng
‘sweetheart’ to address their men, something that was considered improper for
women in the past.
To summarize, the asymmetry in the relationship between husbands and
wives is partly reflected in the terms men use to refer to their wives. Men can
use certain terms when speaking or referring to their wives. In contrast, wives
are not allowed to make use of these terms when addressing or referring to their
husbands. The privilege of being a man is shown in a familiar folk saying:
(34) Khôn ngoan cũng thße 0àn bà, dau
à rÓang vung
» »
dai
smart still be women although clumsy silly
cũng là 0àn ông.
still be men
‘No matter how smart she is, she is still just a woman; no matter
how clumsy and silly he is, he is still a man.’
310 Hoa Pham
This saying no longer has credence. Wives do not depend economically on their
husbands as before. Today, women usually work for a living. They participate
in decision-making about every matter, including their children’s education.
Husbands and wives have almost equal roles to play in family matters. The
terms they use for address and self-reference, therefore, have changed reflecting
ongoing changes in the relationships between women and men.
6. Conclusion
Terms of address, self-reference and reference have changed in the last few
decades with the change of women’s roles in society. Men have had to change
the ways in which they refer to women, particularly their wives. They can no
longer use terms which assume that men are more privileged than women, such
as the pronouns tao … mày ‘I … thou’, nó ‘she’ or the special designation con.
Women have become more sensitive about the terms they use for self-
reference and for addressing, especially when they speak to men. They have
more choices in different situations, such as using proper names for self-
reference and more intimate terms in referring to husbands or boyfriends.
In other relationships, women also use more neutral and formal terms for
self-reference when speaking to men, such as proper names and the pronoun
tôi. Men, in contrast, employ more courteous and respectful terms to refer to
women, such as ch»i ‘elder sister’ instead of cô ‘father’s sister’.
The context-sensitive characteristics of such terms, in turn, reflect social
changes in the relationship between women and men. I have illustrated some of
these changes through the terms women and men use in various relationships
in urban settings among young or educated people. The changes seem to follow
the trend toward more symmetric expression. Women are referred to by more
polite expressions. They are allowed to express their feelings more freely by
using relaxed and intimate terms to address men.
While the system provides a rich set of terms that is very context-sensitive
and is also able to accommodate social changes, men and women in certain
situations nevertheless find themselves confronted with some tricky problems.
For example, adult male students must determine how to address younger
female teachers, and older women must decide how to refer to themselves when
speaking to their younger husbands, or educated young women must find
suitable terms to refer to themselves when speaking to lesser educated men.
These are all situations which arose very rarely, if at all, in Vietnamese society
<DEST "pha-n*">
Vietnamese 311
before about 1950. Now these situations are not uncommon, and they are
growing more common.
The traditional Vietnamese system of address terms was established, of
course, over many generations of relatively static social relationships. With the
increasing fluidity of social relationships in the second half of the 20th century,
a new system of address and self-reference is necessarily evolving. One useful
stratagy is to avoid using traditional terms altogether. Another, which is more
effective and has become widespread, is to use proper names. This change itself
appears to be symbolic. Proper names help to characterize a person – to some
extent, they identify a person as an individual. Perhaps the social changes in
Vietnam reflect the loosening of traditional categories, in which people were
considered members of groups with well-defined and invariant niches in the
social order. The new order is much more flexible, and the changes we see in
the use of terms of address, self-reference and reference are a clear indicator of
social change.
Notes
* I would like to thank Jack Chambers for his advice, encouragement and strong support,
and the editors for their careful editing and suggestions. Any errors, of course, are my own.
à (1987, 1997); reference
1. For a general overview of the Vietnamese language cf. Nguyen
grammars are Emeneau (1951), Thompson (1965); a standard Vietnamese-English dictio-
nary is Bùi (1995).
2. (w) is a labio-velar glide; it is the only segment which can occur in the position between
initial consonant and vowel.
3. The two terms 0þc » ‘male’ and cái ‘female’ are used for animals, e.g., sþtþÑ 0þc
» ‘lion’ vs. sþtþÑ
cái ‘lioness’, chó 0þc
» ‘male dog’ vs. chó cái ‘female dog’. However, these terms occur in some
abuse forms for human beings to show disapproval, e.g., 0ı̃ 0þc » ‘a male prostitute’, lai » cái
return + female ‘a gay’.
4. In North Vietnam, children use con for self-reference only with their parents, and cháu
‘nephew, niece, grandchild’ with the others.
5. These words are Sino-Vietnamese and usually occur in compounds with other Sino-
Vietnamese words. The Vietnamese words for the prepositions ‘inside’ and ‘outside’ are trong
and ngoài, respectively. The term for addressing a great-grandparent irrespective of gender
is cu» in Northern dialects and câo in Southern dialects. The term referring to the great-
grandchild is chÔat.
6. Some grammars claim that there are no personal pronouns in Vietnamese, because Vietnam-
ese does not have words that can be used in all contexts with the meanings ‘I, you, he, she, we,
they’ as in English, and because some words with pronominal functions originate from nouns.
</TARGET "pha">
312 Hoa Pham
References
»
Bùi, Phung. 1995. Vietnamese-English dictionary. Hanoi: The Gioi.
Ðào, Duy Anh. 1973. Chþ̃ Nôm: Nguáon gâoc, câau tao, » diàen biâen [Nom language: Origin,
» Khoa Hoc
structure, process]. Hà Nôi: » Xã Hôi. »
Dodd, Jan & Mark Lewis. 1998. Vietnam: The rough guide. London: Rough Guides.
Emeneau, M. B. 1951. Studies in Vietnamese (Annamese) grammar. Berkeley: University of
California Publications in Linguistics.
Khái Hþng. ca 1930. Cái Ve [A girl named Ve]. Sài gòn: Song â Mǘi.
Löbel, Elisabeth. 2000. “Classifiers versus genders and noun classes: A case study in Vietnam-
ese.” In Gender in grammar and cognition. eds. Barbara Unterbeck et al., Berlin: de
Gruyter, 259–319.
Lþüng, Hy V. 1990. Discursive practices and linguistic meanings. The Vietnamese system of
person reference. Amsterdam: Benjamins.
»
Manh, Phú Tþ. 1939. Làm lẽ [Being a concubine]. Sài gòn: Xuân Thu. (Reprint 1967).
Ngô, Thanh Nhàn. 1984. The syllabeme and pattern of word formation in Vietnamese. Ph.D.
dissertation. New York University.
à Tài Can.
Nguyen, ß 1977. Ngþ̃ pháp tiâeng Viêt:» tiâeng – tþ̀ ghép – 0oan
Ñ ngþ̀ [Vietnamese gram-
mar: Word – compound – phrase]. Hà nôi: » Ðai» hoc » Quocâ Gia Hà nôi.
»
à Ðình Hòa. 1987. “Vietnamese.” In The world’s major languages, ed. Bernard
Nguyen,
Comrie. London: Croom Helm, 777–796.
à Ðình Hòa. 1997. Vietnamese. Amsterdam: Benjamins.
Nguyen,
Thompson, Laurence C. 1965. A Vietnamese grammar. Seattle: University of Washington Press.
Vũ, Ngoc» Phan. 1994. Tuc » ngþ̃, ca dao, dân ca Viêtnam» [Vietnamese proverbs and folk
» Khoa Hoc
songs]. Hà nôi: » Xã Hôi » Press.
<TARGET "awb" DOCINFO
AUTHOR "Gwenllian Awbery, Kathryn Jones and Delyth Morris"
TITLE "The politics of language and gender in Wales"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
welsh
The politics of language and gender in Wales
Gwenllian Awbery
University of Wales, Cardiff, UK
Kathryn Jones
Lancaster University, UK
Delyth Morris
University of Wales, Bangor, UK
1. Introduction
2. Welsh versus English: Language shift and the struggle for legitimacy in
the public domain
3. Selected structural properties of the Welsh language
3.1 Grammatical gender and referential gender
3.2 The pronoun system
4. Bringing the language and gender debate to centre stage
5. Guidelines for non-discriminatory language use
6. The ongoing debate
Notes
References
1. Introduction
Welsh (Cymraeg) is an Indo-European language, a member of the Celtic

subgroup. It is closely related to Breton and Cornish, and more distantly to
Irish, Scottish Gaelic and Manx. It is currently spoken by approx. 500,000
speakers, i.e. 20% of the population of Wales, one of the countries which make
up the British Isles. It has been under heavy pressure from English since the
16th century, but since the 1960s a growing demand for language rights has
resulted in greater official status for the Welsh language, and it is now used in
a wider range of domains than previously.
314 Gwenllian Awbery, Kathryn Jones and Delyth Morris
Welsh is a VSO language, adjectives follow a head noun, and there are
prepositions rather than postpositions. It has two grammatical genders,
traditionally referred to as masculine and feminine, but does not mark gram-
matical case. Verbs agree with a pronominal subject in number and person, but
all nominal subjects require a third person singular verbal inflection. Preposi-
tions agree with a following pronominal object in number, person and (in the
third person singular) in gender, but show no agreement with a nominal object.
Like all Celtic languages, Welsh has a complex system of Initial Mutation,
whereby the initial phonological segment of a word is changed in certain lexical
and grammatical environments.1
It is important to understand issues of gender in relation to the Welsh
language within the context of the politics of marginalised/minority languages.
Welsh speakers constitute a language group that is also a social group, with its
own markers of language and culture, and one which exists within a state where
it lacks the political and ideological structures necessary to guarantee the
relevance of the language to the everyday life of its members (European
Commission 1996).
The emergence of the rational modern state in the 18th century was
accompanied by an increasing rationalism in government and administration,
together with a surge towards scientific and technological progress. In the UK,
English, as the powerful state language, became identified with modernisation
and progress, while Welsh, along with other autochthonous minority languages
such as Scottish Gaelic, was deemed to be traditional and non-progressive, and
by definition, therefore, unable to grapple with the demands of modern life.
English consequently became acknowledged as the appropriate language for
work, administration, education, science, and technology, while Welsh retreat-
ed into civil society, becoming the language of the family, religion, and commu-
nity (Williams 1992).
From a gender aspect, the link that is often made between rationality/
science and the male and, by contrast, irrationality/emotion and the female, is
a further consideration. The subordination and marginalisation of women has
been the prime struggle of the feminist movement, and similarly the subordina-
tion of the minority language has been the major issue for the Welsh language
movement in Wales. The overlap between gender inequality and language
inequality, and the linking of male/English/rationality and female/Welsh/
irrationality, means that one type of inequality reinforces the other (Morris &
Williams 1994: 132), and it is this particular combination, with its resultant
discriminatory effects, that led the Equal Opportunities Commission to
Welsh 315
undertake research into issues of gender and the Welsh language in Wales
(Awbery 1997, Jones & Morris 1997).
This chapter locates the recent attention paid to non-sexist language use in
Welsh within the context of Welsh language politics. Section 2 outlines the
history of the struggle for the survival of Welsh, in particular the campaigns of
the Welsh Language Society, which had priority for most Welsh speaking
feminists over issues about language and gender equality, as these issues were
being debated with regard to English. The third section outlines the structural
properties of Welsh. In the fourth section we explain how recent legislation
regarding the use of Welsh in public sector institutions has brought the use of
Welsh for job advertisements to the attention of the Equal Opportunities
Commission (EOC). This section is followed by a summary of the EOC’s
guidelines for non-discriminatory language use in Welsh job advertisements.
Finally, we conclude this chapter with a brief account of the way the guidelines
were disseminated and the reactions of the media and professionals in the field
to the recommendations.
2. Welsh versus English: Language shift and the struggle

for legitimacy in the public domain
In Wales, the discourse of language politics centres around the power struggle
between Welsh and English, and the fight for the survival of Welsh as a viable
language of social life in the face of socio-economic globalisation, and the
pervasiveness of English in the public domain. In quantitative terms, succes-
sive Census of Population surveys have shown that, during the course of the
century, the number of so-called Welsh speakers in Wales has almost halved,
from 922,100 in 1921 to 508,000 by 1991 (OPCS 1994). Recently, there are
signs of growth in the number of Welsh speakers amongst younger age
categories within the population (Williams 1989: 44, Davies 1993: 67–72).
However, the percentage of Welsh speakers in Wales as a whole deteriorated
markedly between 1921 and 1991, with 37 per cent of the total population
speaking Welsh in 1921, compared to 19 per cent seventy years later. This shift
may be attributed partly to immense socio-economic changes during the
course of the century which have led to massive inward and outward migration
– the outward migration of Welsh speakers and inward migration of monoglot
English speakers (cf. Osmond 1987). Other pertinent factors in the decline of
the number and proportion of speakers are the two world wars, when thou-
sands of young male Welsh speakers were killed, and increasing secularisation,
which has meant that the traditional nonconformist chapels have lost their role
as reproducers of the Welsh language (Williams 1987, Price 1994). Similar
problems have befallen other minority languages in Europe, like Basque and
Frisian (Gardner n.d., van Langevelde 1994), as the usual agents of language
transmission, particularly the community and family, have been eroded by
changes in the economic structure. In the face of such an alarming decrease in
the number of Welsh speakers and the consequent lack of opportunities for
using Welsh as a language of social interaction, efforts to restrain the momen-
tum of language shift have focused upon challenging the social and institution-
al hegemony of English.
English had gained domination over Welsh as the legitimate language of
official life by the end of the 15th century (Davies 1993). This situation was
cemented in 1536 by the legislation which unified Wales with England when the
Act of Union stipulated that English, not Welsh, should be the only language of
the courts and public office. The legitimacy of Welsh as a language of education
was challenged when the Report of the Commissioners of Inquiry into the State of
Education in Wales of 1847, popularly known as the Blue Books, vigorously
condemned the use of Welsh. As a consequence of this report and the Educa-
tion Act of 1870, a predominantly English-medium education system became
established throughout Wales. Efforts to counteract the negative attitude
towards Welsh instilled by the “Treachery” of the Blue Books, most notably the
report Welsh in Education and Life (H. M. S. O. 1927) concentrated on re-
legitimising Welsh as a language for education. The report called for increased
resources to be allocated to training teachers and for the preparation of appro-
priate teaching materials through the medium of Welsh. By the mid-1940s,
Welsh had become the main teaching medium in schools in the more Welsh-
speaking areas, and attempts were made to teach some Welsh to monoglot
English speakers in the more Anglicised parts of Wales. The rapid growth of
Welsh-medium schools has perhaps been one of the most remarkable develop-
ments in recent years, from the first state school designated as Welsh-medium
in Llanelli in 1947 to 444 Welsh-medium primary schools and some 52 Welsh-
medium secondary schools by 2000 (The National Assembly for Wales 2001:
Table 7.8, p. 49).
Efforts to re-establish Welsh as a legitimate language in public life have
principally taken the form of campaigns organised by various social movements
and pressure groups. Undeb y Cymdeithasau Cymraeg (The Union of Welsh
Societies) was one such group which was responsible for galvanising public
Welsh 317
support and organising a petition following the trial and subsequent imprison-
ment in 1936 of three leading Welsh nationalists, Saunders Lewis, Lewis
Valentine and D. J. Williams. The furore that accompanied the judge’s refusal
to allow the defendants to address the court in Welsh encouraged three-
quarters of a million people in Wales to sign a petition demanding that Welsh
be granted equal status with English. Although this was not achieved, the Welsh
Courts Act of 1942 noted that “the Welsh language may be used in any court in
Wales by any party or witness who considers that he [sic] would otherwise be at
a disadvantage by reason of his natural language of communication being
Welsh” (Davies 1993: 64).
Despite this legislation, by the middle of the century Welsh held virtually
no status as a legitimate language in Welsh public life. Saunders Lewis’
renowned BBC broadcast on the “Fate of the Language” in 1962 proclaimed
the need to employ revolutionary means in order to prevent the demise of the
Welsh language and exhorted his audience to “make it impossible to conduct
local authority or central government business in Wales without the Welsh
language” (Davies 1993: 94). This broadcast provided the impetus for the
founding of Cymdeithas yr Iaith Gymraeg (The Welsh Language Society). Since
its foundation in the summer of 1962, Cymdeithas yr Iaith members and
supporters have conducted numerous campaigns involving non-violent direct
action that regularly lead to police action, prosecution, fines and imprison-
ment. Their early campaigns concentrated on targeting public sector institu-
tions to secure, for example, printed forms in Welsh and bilingual roadsigns
throughout Wales. It was largely Cymdeithas yr Iaith’s members’ and support-
ers’ refusal to pay monolingual English bills and their demand for printed
forms in Welsh which galvanised the public pressure that eventually led to the
passing of the Welsh Language Act of 1967. This Act further improved the
status of Welsh in the courts and with regard to official and public administra-
tion (Clowes 1991). The 1967 Act adopted the principle of “equal validity”,
whereby any act, documents or anything else done through the medium of
Welsh in Wales would have the same legal force as it would in English.
However, although the 1967 Act gave Ministers of the Crown the power to
provide Welsh and bilingual documents, they were not legally obliged to do so
and consequently in practice little change took place.
Campaigning for the right to use Welsh more extensively within the public
domain continued during the 1970s and 1980s. One of the major campaigns
was for a Welsh television channel which was finally established in 1982. The
extended use of Welsh in radio and television broadcasting has, on the one
hand, been credited with evolving a “suppler, more popular speech” (Davies
1993: 63) and on the other hand, with undermining the accepted and recog-
nised syntax of the Welsh language by introducing English syntax and idioms
into general speech patterns. The ineffective implementation of the 1967 Act’s
provisions also led to further campaigning, which eventually culminated in the
most recent language legislation, the 1993 Welsh Language Act. This Act
established a Welsh Language Board charged with the function of encouraging
all public sector bodies in Wales to use Welsh on a basis of equality with English
in all their dealings with the public. In spite of its limited powers, the 1993
Language Act has significant implications as a means of finally effecting some
change in the status and use of Welsh in the public sector.
Given the focus on linguistic survival, the language and gender debate has
played a relatively marginal role in Welsh language politics. In the 1970s, when
feminists were challenging how women were represented and positioned
linguistically in English, most Welsh feminists were more concerned with
supporting Cymdeithas yr Iaith’s language campaigns, than fuelling public
controversy about sexist language use in Welsh. This is not to say that Welsh
speakers have not been sensitive to the representation of social gender identities
in Welsh, but to emphasise the lack of prominence given to this issue publicly.
3. Selected structural properties of the Welsh language
Welsh, like many other languages, divides nouns into two sets, which behave
differently with respect to various grammatical rules. To some extent this
distinction corresponds to the real world difference between male and female,
with nouns which refer to males being found in one set, and nouns which refer
to females being found in the other. The two sets are traditionally said to be of
masculine and feminine grammatical gender.
3.1 Grammatical gender and referential gender

The real world difference between male and female does not, however, align
neatly with grammatical gender in every case, and there are a number of
asymmetries. Where a noun refers specifically to a man or a woman (due to its
lexical gender), its grammatical gender is normally masculine or feminine as
appropriate: tad (m) ‘father’, bachgen (m) ‘boy’, brawd (m) ‘brother’, mam (f)
‘mother’, merch (f) ‘girl’, chwaer (f) ‘sister’. There are, however, many other
Welsh 319
nouns without lexical gender which are grammatically masculine, but are used
freely to refer to either a man or a woman: plentyn (m) ‘child’, cyfaill (m)
‘friend’, meddyg (m) ‘doctor’. Occasionally, and for the most part in a rather old
fashioned style, a feminine suffix can be added to such forms to give a specifi-
cally feminine form, such as cyfeilles (f) ‘female friend’. This is not possible in all
cases, and there is for instance no such form as *plentynes ‘female child’.
Feminine forms such as these should be considered as marginal in the modern
language. Forms such as plentyn, cyfaill and meddyg which can refer freely to
both male and female referents may therefore be best referred to as masculine
epicenes.2 In contrast, there are very few nouns which are grammatically of
feminine gender and may be used to refer freely to either a man or a woman:
cennad (f) ‘messenger’ and nyrs (f) ‘nurse’.
A similar picture emerges in the case of animals, birds and other creatures,
which display gender differences. Where a noun refers specifically to a male or
to a female, its grammatical gender is normally masculine or feminine as
appropriate: tarw (m) ‘bull’, march (m) ‘stallion’, ceiliog (m) ‘cockerel’, buwch
(f) ‘cow’, caseg (f) ‘mare’, iâr (f) ‘hen’. A noun referring generically to a species,
on the other hand, is arbitrarily of masculine or feminine gender: ci (m) ‘dog’,
crwban (m) ‘tortoise’, eryr (m) ‘eagle’, cath (f) ‘cat’, llygoden (f) ‘mouse’,
gwennol (f) ‘swallow’. The distinction between masculine and feminine gram-
matical gender also extends to nouns which refer to inanimate objects and
abstract concepts, where again the choice of gender is arbitrary: mynydd (m)
‘mountain’, llyfr (m) ‘ book’, gobaith (m) ‘hope’, afon (f) ‘river’, tref (f) ‘town’,
addysg (f) ‘education’.
It is not possible, in most cases, to tell whether a noun is grammatically
masculine or feminine in gender purely from its form. There is, for instance,
nothing to mark taid (m) ‘grandfather’ or golau (m) ‘light’ as masculine in
gender, and nothing to mark nain (f) ‘grandmother’ or gardd (f) ‘garden’ as
feminine. There are some suffixes which are used to derive specifically mascu-
line or feminine nouns, but they provide a clue only for a limited number of
nouns. The suffix -ydd, for instance, is always masculine, as in cyfieith-ydd (m)
‘translator’, while -es is always feminine, as in actor-es (f) ‘actress’. There are
also some instances of paired suffixes, one masculine and the other feminine.
Pairs of this kind are -yn/-en, as in coch-yn (m) ‘a red-headed man’/coch-en (f)
‘a red-headed woman’ and -wr/-wraig, as in cyfarwydd-wr (m) ‘male director’/
cyfarwydd-wraig (f) ‘female director’, though there is also evidence of such
apparently masculine forms as cyfarwyddwr being used as a masculine epicene,
referring equally to females.
A more certain indicator of the gender of a noun is its behaviour with

respect to certain agreement rules, among them those which concern a noun
following the definite article, and an adjective following a noun. Where the
definite article precedes a noun, the form of the noun varies according to
whether it is masculine or feminine in grammatical gender. A masculine noun
such as bachgen (m) ‘boy’ retains its citation form, to give y bachgen ‘the boy’.
A feminine noun such as merch (f) ‘girl’ is subject to an Initial Mutation Rule,
which changes its initial consonant to give y ferch ‘the girl’. Similarly, an
adjective following a masculine noun retains its citation form, but an adjective
following a feminine noun undergoes an Initial Mutation Rule. The adjective
bach ‘little’ is therefore unchanged in the phrase bachgen bach ‘a little boy’, but
has an altered initial consonant in the phrase merch fach ‘a little girl’.
This distinction between masculine and feminine gender is, however,
relevant here only in the singular. In the plural there are no grammatical
distinctions dependent on gender, though the semantic distinctions still hold.
The definite article is followed by the citation form of both masculine and
feminine nouns, as y bechgyn (m) ‘the boys’ and y merched (f) ‘the girls’. The
adjective, too, appears in citation form following both masculine and feminine
nouns, as in bechgyn bach ‘little boys’ and merched bach ‘little girls’.
3.2 The pronoun system

There are distinct third person singular masculine and feminine pronoun
forms in Welsh, used to refer to an individual man or woman, i.e. e, fe (m) ‘he’,
‘him’ and hi (f) ‘she, her’. The distinction is obligatory, and shows up in all
positions in the sentence. Examples (1) and (2) illustrate pronouns in subject
function, with the subject following the verb, while (3) and (4) show pronouns
in object function.
(1) Mae e ’n mynd.
is he prog go.inf
‘He is going.’
(2) Mae hi ’n mynd.
is she prog go.inf
‘She is going.’
(3) Clywais i fe.
heard I him
‘I heard him.’
Welsh 321
(4) Clywais i hi.

heard I her
‘I heard her.’
Where the pronoun follows a preposition, there is a distinct inflection on the

preposition, agreeing with the masculine and feminine pronoun. In (5) and (6),
for instance, the preposition am ‘for’ is inflected to agree with the following
pronoun. Here the pronoun is retained in the sentence, but it could equally well
be dropped, leaving the inflection as the only overt marker.
(5) Soniais i amdano fe.
spoke I about.3sg.masc him
‘I spoke about him.’
(6) Soniais i amdani hi.
spoke I about.3sg.fem her
‘I spoke about her.’
No such inflections appear where a preposition is followed by a noun; in such

forms the preposition is unaffected by the gender of the following noun, as in
am y bachgen ‘about the boy’, and am y ferch ‘about the girl’.
Masculine and feminine possessive pronouns are identical in form, but
nevertheless trigger different Initial Mutation Rules in the following “possessed”
noun. The noun cath (f) ‘cat’, for instance, appears as gath when it follows the
masculine possessive pronoun, ei gath ‘ his cat’, but as chath when it follows the
feminine possessive pronoun, ei chath ‘her cat’.
The distinction between masculine and feminine pronoun forms, however,
is limited to the third person singular. In the plural there is only one set of
forms, which do not distinguish referential gender. ‘They’ may be a group of
men, a group of women, or a mixed group.
(7) Maen nhw yn mynd.
are they prog go.inf
‘They are going.’
(8) Clywais i nhw.
heard I them
‘I heard them.’
(9) Soniais i amdanyn nhw.
spoke I about.3pl them
‘I spoke about them.’
The plural possessive pronoun does not trigger an Initial Mutation Rule, and
the following noun retains its citation form following this possessive, eu cath
‘their cat’. Similarly, first and second person pronouns do not specify referential
gender, either in the singular or the plural. Gender is simply not a relevant
feature except in the third person singular, where it is obligatorily specified in
the pronoun system.
Since pronouns are chosen according to the gender of the individual
referred to, the masculine form for a man and the feminine form for a woman,
there are many straightforward cases. In (10), a male is referred to by a noun of
masculine gender and subsequently by a masculine pronoun:
(10) Soniodd y bachgen am ei gath.
spoke the boy about his cat
‘The boy spoke about his cat.’
Similarly in (11), a female is referred to by a noun of feminine gender and

subsequently by a feminine pronoun:
(11) Soniodd y ferch am ei chath.
spoke the girl about her cat
‘The girl spoke about her cat.’

There is no problem either if the noun is one of those which may refer freely to
either a male or a female, and its grammatical gender happens to be compatible
with the referential gender. In almost all such cases the grammatical gender of
the noun will be masculine, and referring to a male is therefore unproblematic,
as we have again a masculine noun and then a masculine pronoun:
(12) Soniodd y plentyn am ei gath.
spoke the child about his cat
‘The child spoke about his cat.’
The possessive pronoun ei here is masculine and the appropriate Initial Muta-
tion Rule changes the initial consonant of the following word to give gath,
rather than cath, as noted earlier in Section 3.2, and as seen in (10). A noun
such as plentyn (m) ‘child’, however, may equally well refer to a female child,
and in this case there is a switch, in that a noun of masculine gender is followed
by a feminine pronoun:
Welsh 323
(13) Soniodd y plentyn am ei chath.

spoke the child about her cat
‘The child spoke about her cat.’
Here the Initial Mutation Rule requires a different change in the following
noun, and we find chath, rather than gath, as in (11) above. Such forms may
appear odd, but the need to refer to an individual person by the referentially
appropriate pronoun appears strong enough to override the presence, in the
same phrase or sentence, of a noun of the opposite grammatical gender. The
crucial consideration is the gender of the individual referred to rather than the
arbitrary categorisation of the noun within the language.
Pronouns are, it appears then, sensitive to referential gender. Other rules
of the grammar are not, and make no distinction between a masculine noun
referring to a male individual, and a generic masculine which may refer to
either a male or a female. A noun such as plentyn (m) ‘child’, which may refer
to either a boy or a girl, behaves with reference to these other rules in the same
way as a straightforwardly masculine noun such as bachgen (m) ‘boy’. Both
retain their citation form following the definite article, as in y plentyn ‘the
child’ and y bachgen ‘the boy’. In both cases an adjective following the noun
appears in citation form, as bach ‘little’ in plentyn bach ‘little child’ and bachgen
bach ‘little boy’.
4. Bringing the language and gender debate to centre stage
For more than twenty years in the UK, the Sex Discrimination Act has made it
unlawful to treat anyone less favourably than anyone else on the grounds of his
or her gender. This applies to employment, education and training, and the
provision of goods, facilities and services to the public. The deep structural
inequalities that remain indicate that the process has not been altogether
successful, partly because the Act itself was not specific enough in many areas of
gender inequalities. One area where the Act was specific, however, was in the
field of job advertisements, where it stated:
It is unlawful to publish or cause to be published an advertisement which
indicates, or might reasonably be understood as indicating, an intention by a
person to do any act which is or might be unlawful […] use of a job descrip-
tion with a sexual connotation […] shall be taken to indicate an intention to
discriminate, unless the advertisement contains an indication to the contrary.
(H. M. S. O. 1975: Section 38, Parts 1 & 3)
The Equal Opportunities Commission was set up to oversee the working of the
Act, and job advertisements is one sphere where it has been comparatively
successful in changing discriminatory practices.
In English, the kind of job titles which could indicate an intention to
discriminate would be barman or salesgirl, and unless the advertisement
contained a disclaimer to the contrary, the organisation could be liable to
prosecution in the courts of law. For a number of years, the Equal Opportuni-
ties Commission has given guidelines to employers as to good practice in
advertising jobs in English, but until recently, no such guidelines were available
in Welsh. The demands of the 1993 Welsh Language Act, which makes it
necessary for public bodies in Wales to publish language plans outlining their
Welsh-medium service provision to those customers who required such a
service, meant that there would be an increasing use of Welsh in job advertise-
ments, as organisations which had previously not used Welsh in work were
obliged to fill certain posts with Welsh/English bilingual personnel. The Equal
Opportunities Commission in Wales therefore commissioned research that
would enable it to formulate guidelines for employers advertising (perhaps for
the first time) in Welsh, to ensure that they fulfilled their legal obligations, and
to ensure that Welsh-speaking women and men were not deterred from
applying for certain types of employment because of the inappropriate use of
job titles. The publication of The Sex Discrimination Act and the Use of Welsh in
the Workplace (Awbery 1997) marked the beginning of the Commission’s
process of public consultation within Wales with employers, linguists, academ-
ics, and other interested parties, with a view to gaining a consensus on the
correct and proper way to advertise jobs in Welsh in the future. Further
research on Gender and the Welsh Language (Jones & Morris 1997) was com-
missioned by the Equal Opportunities Commission in Wales to consider the
more wide-ranging impact of the Act on the lives of Welsh-speaking and non-
Welsh speaking women and men in Wales.
5. Guidelines for non-discriminatory language use
The Equal Opportunities Commission’s duty to ensure that job advertisements

in Welsh, as in English, should not appear to discriminate against men or
women in an unlawful manner means that it has a responsibility to monitor job
advertisements and ensure that they are worded in a nondiscriminatory way.
Over the years, considerable agreement has been reached as to what counts as
Welsh 325
non-discriminatory usage in English, but in Welsh the situation was until

recently much less clear (Awbery 1997). Furthermore, at the time the guidelines
were commissioned, while some employers took care to use non-discriminatory
language in both Welsh and English in their advertisements (as in Figure 1),
others did not.
Job advertisements usually include very prominently a single noun,
specifying the staff needed to fill the vacant post. It is important that this noun
make it clear that both men and women are eligible to apply and that there is no
intention to discriminate. Many job titles used in Welsh fall into the set of
nouns which, though grammatically of masculine gender, may refer freely to
men and women, and it is recommended that these be used whenever possible:
trefnydd (m) ‘organiser’, cyfrifydd (m) ‘accountant’.
Figure 1.
In other cases the relevant nouns are gender-specific, and it is necessary to

use two forms side by side to ensure that both men and women are clearly
included: athro (m) ‘male teacher’, athrawes (f) ‘female teacher’, newyddiadurwr
(m) ‘male journalist’, newyddiadurwraig (f) ‘female journalist’. Masculine forms
such as newyddiadurwr are occasionally used to refer to both men and women,
but in the case of job advertisements it was felt that there was an overriding
need to ensure that it was clear that the post was open to both men and women.
The suffixes used in such pairs are lexically specific, but only certain lexical pairs
are morphologically marked in this way.
In some cases a job has until recently been carried out only by men or only
by women, and the term is normally interpreted as referring to the group
traditionally associated with that job. If the noun concerned is not explicitly
marked morphologically as referring to one gender, it may be used, but with a
clear statement that the job is in fact open to both females and males. The
creation of new forms to fill gaps, and to ensure that there are clearly male and
female forms available is something that should only be done with great care, as
these new forms can often strike people as odd, and cause unnecessary hostility
to the principle of non-discriminatory wording.
Where there are lexical gaps, with no existing noun available to refer to a
man or woman in a nontraditional job, it is recommended that alternative
phrasing be used, which reads naturally in the language and does not strike the
reader as odd. Instead of searching for an acceptable male equivalent to
ysgrifenyddes (f) ‘female secretary’, for instance, one can use a phrasal equivalent
not marked for gender: swydd ysgrifenyddol (f) ‘secretarial post’. This alternative
phrasing also enables one to avoid the duplication which arises where there is
a need to use two forms side by side, one referring to a man and the other
referring to a woman, as with the paired forms athro (m) ‘male teacher’ and
athrawes (f) ‘female teacher’: swydd ddysgu (f) ‘teaching post’.
Another approach is to use the plural form, with a slightly reworded
advertisement, since this can refer to men and women equally:
(14) Gwahoddir ceisiadau gan athrawon profiadol.
is.invited applications from teachers experienced
‘Applications are invited from experienced teachers.’
The wording of the rest of the advertisement is also important. Here the
qualities, qualifications and experience required of the applicant are spelled out,
and here again it is important that the wording makes it clear that both men
and women are eligible to apply, with general terms, which can refer freely to
Welsh 327
both men and women, being used to refer to applicants: ymgeisydd (m) ‘can-
didate’, deilydd y swydd (m) ‘postholder’.
Where the advertisement is framed in the singular, the male and female
pronouns will need to be used side by side in order to ensure that it is clearly
seen to refer to both men and women:
(15) Bydd yr ymgeisydd llwyddiannus yn […]
will.be the candidate successful prog
Rhaid iddo ef/iddi hi feddu
need for.3sg.masc him/for.3sg.fem her possess.inf
ar gymwysterau da.
on qualifications good
‘The successful candidate will be […] It is necessary for him/for her to
possess good qualifications.’
This duplication can be avoided by the use of plural forms, where the pronoun
system does not distinguish between male and female:
(16) Dylai ymgeiswyr fod yn barod i […]
should candidates be.inf prt ready to
Rhaid iddynt feddu ar gymwysterau da.
need for.3pl possess.inf on qualifications good
‘Candidates should be ready to […] It is necessary for them to possess
good qualifications.’
Alternatively the requirements can be expressed in the second person, where

once again the pronoun system does not distinguish between male and female:
(17) Byddwch yn gweithio […]
will.be.2pl prog work.inf
Rhaid i chi feddu ar gymwysterau da.
Need for you.2pl possess.inf on qualifications good
‘You will be working […] You will need to possess good qualifications.’
It is also possible to express the needs of the post in ways which do not refer to
the individual at all, and therefore do not carry any implications as to the
gender of the person involved:
(18) Mae profiad blaenorol yn hanfodol.
is experience previous prt essential
‘Previous experience is essential.’
(19) Bydd angen gweithio oriau hyblyg.

will.be need work.inf hours flexible
‘It will be necessary to work flexible hours.’
These guidelines put forward by the Equal Opportunities Commission are

therefore formulated largely in terms of a “neutralisation” strategy. Job adver-
tisements are formulated so as to omit reference to the gender of the candidate
wherever possible. The alternative “visibility” strategy, which would involve
creating forms to give male/female pairs was avoided, except where it is
necessary to have masculine and feminine pronouns side by side as in (15), as
it was felt that the Welsh-speaking community would regard the introduction
of new, nonstandard lexical forms as linguistic imperialism on the part of
English-speaking officials, with no understanding of or sympathy with the
Welsh language. It was felt to be important that the guidelines should recom-
mend forms which were indisputably correct Welsh, and that the issue of
ensuring that job advertisements are framed in a non-discriminatory way
should not become entangled in resentment of interference from English.
6. The ongoing debate
The Equal Opportunities Commission consulted widely over a period of six

months, to establish whether the proposed guidelines were generally acceptable.
Copies of the report were sent to interested parties, such as personnel officers,
those responsible for job advertisements in the press, and translators, and a
one-day seminar was held to explain the thinking behind the guidelines and to
encourage discussion. There was also considerable coverage in the press and on
radio and television.
There was some hostility in the press, in part displaying the usual dismissive
reactions to anything perceived as “feminist”, but also, interestingly enough,
condemning the guidelines as an English imposition, designed to undermine
the natural structure of the Welsh language. In fact, the guidelines were
formulated throughout with the intention that the forms recommended should
be natural to the language, using the flexibility that is already present in normal
usage rather than trying to impose something unfamiliar which might be
perceived as alien and English. The importance of this factor in a minority
language situation was recognised early on and very consciously taken into
account in drawing up the recommendations.
Welsh 329
The reactions of professionals in the field, such as personnel officers and

translators, generally were favourable. They appeared glad to have clear guide-
lines, and felt that on the whole the Commission had struck the right note.
There were some disagreements of detail as to the most appropriate job titles to
use in certain cases, and discussions of the relationship between grammatical
and referential gender in Welsh (Prys 1997). There were no major disagree-
ments of principle, however, and the guidelines suggested by the Equal Oppor-
tunities Commission have now been adopted as official policy.
The final step in this process has been the issuing of a leaflet summarising
the guidelines in a simple, easy to use format, for personnel officers, translators
and others involved in drawing up job advertisements. This leaflet has been
published with the support of the Welsh Language Board. There were initially
some difficulties in arriving at an agreed form of words, reflecting in part some
of the disagreements which emerged during the public consultation exercise,
but also the rather different legal position of the Equal Opportunities Commis-
sion under the Sex Discrimination Act and the Welsh Language Board under
the Welsh Language Act. It is encouraging to note that the two bodies have been
able to collaborate in ensuring that guidelines of this kind are available to those
seeking to extend the use of Welsh in all aspects of public life.
Notes
1. Accessible grammars of Welsh are King (1993) and Thorne (1993). Griffiths & Jones
(1995) is a very good dictionary for English to Welsh; for Welsh to English the University of
Wales Dictionary (Geiriadur Prifysgol Cymru) is good but still incomplete; the most useful
dictionary for everyday purposes is Thorne (1991), which translates in both directions.
2. We are grateful to the Benjamins reader for drawing our attention to some of the
complications found with nouns of this type.
References
Awbery, Gwenllian. 1997. The sex discrimination act and the use of Welsh in the workplace.
Cardiff: Equal Opportunities Commission.
Bevan, Gareth & Patrick Donovan, eds. 1950f. Geiriadur Prifysgol Cymru [University of
Wales Dictionary]. Cardiff: University of Wales Press.
Clowes, Carl. 1991. Language strategy 1991–2001. Caernarfon, Gwynedd: National Lan-
guage Forum.
</TARGET "awb">
Davies, Janet. 1993. The Welsh language. Cardiff: University of Wales Press.
Equal Opportunities Commission. 1999. Advertising jobs in Welsh: Guidelines / Hysbysebu
swyddi yn Gymraeg: Canllawiau. Cardiff: Equal Opportunities Commission.
European Commission. 1996. Euromosaic: The production and reproduction of the minority
language groups in the European Union. Brussels: OPEC.
Gardner, Neil. n.d. Goodwill, language planning, and language policies. Carmarthen: Joint
Working Party on Bilingualism in Dyfed.
Griffiths, Bruce & Dafydd Glyn Jones, eds. 1995. The Welsh Academy English-Welsh dictio-
nary: Geiriadur Yr Academi. Cardiff: University of Wales Press.
Her Majesty’s Stationery Office. 1927. Welsh in education and life. London: HMSO.
Her Majesty’s Stationery Office. 1975. The sex discrimination act. London: HMSO.
Jones, Dafydd Glyn. 1995. “Gwyr, Gwragedd, a Phyncrocwragedd” [Men, women, and punk
rockers]. Barn 386: 48–49.
Jones, Kathryn & Delyth Morris. 1997. Gender and the Welsh language. Cardiff: Equal
Opportunities Commission.
King, Gareth. 1993. Modern Welsh: A comprehensive grammar. London: Routledge.
Morris, Delyth & Glyn Williams. 1994. “Language and social work practice: The Welsh case.”
In Social work and the Welsh language, eds. Rhian Huws Williams & Hywel Williams &
Elaine Davies. Cardiff: CCETSW & University of Wales Press, 123–154.
Office of Population and Census Statistics. 1994. 1991 Census/Cyfrifiad. 1991 Welsh/Cymraeg.
London: HMSO.
Osmond, John. 1987. “A million on the move.” Planet 62: 114–118.
Price, Adam, ed. 1994. Quiet revolution? Language, culture and economy in the nineties.
Aberystwyth: Menter a Busnes.
Prys, Delyth. 1997. “Gender and sex in Welsh nouns.” Planet 121: 88–91.
Thorne, David. 1993. A comprehensive Welsh grammar. Oxford: Blackwell.
Thorne, David, ed. 1991. Collins Spurrell Welsh dictionary. London: Harper Collins.
van Langevelde, Ab. 1994. “Language and economy in Friesland.” Tijdschrift voor Econo-
mische en Sociale Geografie 85: 67–77.
Welsh Office. 1995. Statistics of education and training in Wales: Schools. Cardiff: Govern-
ment Statistical Service.
Williams, Colin. 1989. “New domains of the Welsh language.” Contemporary Wales 3: 41–76.
Williams, Glyn. 1987. “Bilingualism, class dialect, and social reproduction. The Sociology of
Welsh.” International Journal of the Sociology of Language 66: 85–98.
Williams, Glyn. 1992. Sociolinguistics: A sociological critique. London: Routledge.
<TARGET "not" DOCINFO
AUTHOR ""
TITLE "Notes on contributors"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
Notes on contributors
Gwenllian Mair Awbery, born in Cardiff, UK. She received her PhD in Linguis-
tics from the University of Cambridge, UK. She is Coordinating Lecturer for
Welsh at the Centre for Lifelong Learning at Cardiff University. Her main areas
of research are: Welsh linguistics (syntax, phonology, dialectology), and
practical issues relating to the use of Welsh in the public domain.
(1976) The syntax of Welsh: A transformational study of the passive. Cambridge: Cambridge
University Press.
(1997) The Sex Discrimination Act and the use of Welsh in the workplace. Cardiff: Equal
(2001) Development of a standard glossary of Welsh terms in the field of equal opportunity.
Cardiff: Equal Opportunities Commission.
Tove Bull, born in Alta, Norway. She is Professor of Nordic Linguistics at the
Faculty of Humanities of the University of Tromsø. She has been Rector of the
University of Tromsø since 1996. Her main areas of research are: Socio-
linguistics, language contact, and language and gender.
(1985) Lesing og barns talemål. Tromsø-studier i språkvitenskap VII [Reading and children’s
language. Tromsø studies in linguistics VII]. Oslo: Novus.
(1995) “Language maintenance and loss in an originally trilingual area in North Norway.”
International Journal of the Sociology of Language 115: 123–143.
(1995) “Language contact leading to language change: The case of Northern Norway.” In
Linguistic change under contact conditions, ed. Jacek Fisiak. Berlin: de Gruyter, 15–34.
Hadumod Bußmann, born in Frankfurt am Main, Germany. She received her

doctoral degree in German Medieval Studies from the University of Munich.
She was Senior Lecturer (Akademische Direktorin) of German Linguistics at the
Institute of German Philology of the University of Munich. Her main areas of
research are: Linguistic terminology, history of women in academia, and
language and gender.
(1995) with Renate Hof, eds. Genus. Zur Geschlechterdifferenz in den Kulturwissenschaften.
Stuttgart: Kröner.
(1996) Routledge dictionary of language and linguistics. London: Routledge.
332 Notes on contributors
(1997) with Hiltrud Häntzschel, eds. Bedrohlich gescheit. Ein Jahrhundert Frauen und
Wissenschaft in Bayern. München: Beck.
Marjorie K. M. Chan, born in Guangdong Province, China. She received her

doctoral degree in Linguistics from the University of Washington, Seattle, USA.
She is Associate Professor of Chinese Linguistics at the Department of East
Asian Languages and Literatures at the Ohio State University, Columbus. Her
main areas of research are: Chinese linguistics (phonology, dialectology,
language and gender, pragmatics of intonation).
(1991) “Contour-tone spreading and tone sandhi in Danyang.” Phonology 8: 237–259.
(1996) “Gender-marked speech in Cantonese: The case of sentence-final particles je and jek.”
Studies in the Linguistic Sciences 26: 1–38.
(to appear) with Peggy W. Y. Wong & Mary E. Beckman. “An autosegmental-metrical
analysis and prosodic annotation conventions for Cantonese.” In Prosodic typology and
transcription: A unified approach, ed. Sun-Ah Jun. Oxford: Oxford University Press.
Mila Engelberg, born in Helsinki, Finland. She is a PhD candidate at the

Department of General Linguistics at the University of Helsinki. Her main areas
of research are: Gender system and androcentricity in Finnish.
(1993) “Maskuliini, feminiini, virkamies?” [Masculine, feminine, civil servant?]. Naistut-
kimus-Kvinnoforskning 4: 39–50.
(1998) “Sukupuolistuneet ammattinimikkeet” [Gendered occupational terms]. Virittäjä 1:
74–92.
(2001) “Ihminen ja naisihminen: Suomen kielen piilomaskuliinisuus” [Humans and female
humans: Covert masculinity in Finnish]. Naistutkimus–Kvinnoforskning 4:23–36.
Charles Ettner, born in Chicago, Illinois, USA. He received his M. A. in

Anthropology from Stanford University, California, where he is a doctoral
candidate in Anthropology. He is a Lecturer at the Department of Anthropolo-
gy at California State University, Fresno. His main areas of research are:
Teaching and learning English in China, and Chinese in the US; minority
groups, ethnicity, identity and education in China; language and culture of the
She peoples of SE China.
(1993) Sexism and the language reforms of the People’s Republic of China: Socialist language
with Chinese characteristics. MA thesis, Stanford University.
Marinel Gerritsen, born in Haarlem, The Netherlands. She received her

doctoral degree in Historical Sociolinguistics from the University of Leiden. She
holds the Christine Mohrmann chair of Intercultural Business Communication
at the Center for Language Studies and the Department of Business Communi-
cation Studies of the University of Nijmegen. Her main areas of research are:
Sociolinguistics and intercultural communication.
(1997) with Hans van der Velde & Roeland van Hout. “Watching Dutch change: A real time
study of variation and change in standard Dutch pronunciation.” Journal of Socio-
linguistics 1: 361–393.
(1999) “Divergence of dialects in a linguistic laboratory near the Belgian-Dutch-German
border: Similar dialects unter the influence of three different standard languages.”
Language Variation and Change 11: 43–67.
(2000) with Hubert Korzilius & Frank van Meurs & Inge Gijsberts. “English in Dutch
commercials: Not understood and not appreciated.” Journal of Advertising Research
40: 17–36.
Anna Gunnarsdotter Grönberg, born in Alingsås, Sweden. She is a PhD

candidate at the Department of Swedish Language at Göteborg University. Her
main areas of research are: Sociolinguistics, social dialectology, youth language,
language and gender, modern Icelandic, and translation studies.
(1995) Otäckstyget och annat otyg: En analys av namn, stilmotsvarigheter och semantiska
motsvarigheter i den isländska översättningen av Astrid Lindgrens Bröderna Lejonhjärta
[An analysis of proper names, and stylistic and semantic equivalence in an Icelandic
translation of a book by Astrid Lindgren]. MISS 6. Göteborg: Institutionen för
svenska språket vid Göteborgs universitet.
(1997) “‘Arvet från Albert och Herbert till Robert’. Om ungdomsspråket i Göteborgs-
regionen” [‘Dialectal heritage’. On youth language in the region of Göteborg]. In
Ungdomsspråk i Norden, eds. Ulla-Britt Kotsinas & Anna-Brita Stenström & Anna-
Malin Karlsson. MINS 43. Stockholm: Institutionen för nordiska språk vid Stock-
holms universitet, 111–120.
(2001) “Ungdomar + dialekt = sant?” [Youth + dialect = true?]. In Jallaspråk, slanguage og
annet ungdomsspråk i Norden [Jalla-language, slanguage and other kinds of youth
language in the Nordic countries], eds. Anna-Brita Stenström & Ulla-Britt Kotsinas
& Eli-Marie Drange. Kristiansand: Høyskoleforlaget.
Hall, Kira, born in Virginia, USA. She received her doctoral degree in Linguis-
tics from the University of California at Berkeley. She is Assistant Professor of
Linguistics and Anthropology at the University of Colorado at Boulder. Her main
areas of research are: Language, gender and sexuality, linguistic anthropology,
narrative and identity.
(1995) with Mary Bucholtz, eds. Gender articulated: Language and the socially constructed self.
New York: Routledge.
(1997) with Anna Livia, eds. Queerly phrased: Language, gender, and sexuality. New York:
Oxford University Press.
(1998) “Lip service on the fantasy lines.” In The feminist critique of language, ed. Deborah
Cameron. New York: Routledge, 321–342.
Marlis Hellinger, born in Pirna/Dresden, Germany. She received her doctoral

degree in English Linguistics from the University of Hamburg. She is Professor
of English Linguistics at the Department of English and American Studies at the
University of Frankfurt am Main. Her main areas of research are: Contrastive
linguistics, sociolinguistics, language and gender, and creole linguistics.
(1985) Englisch-orientierte Pidgin- und Kreolsprachen: Entstehung, Geschichte und sprachlicher
Wandel. Darmstadt: Wissenschaftliche Buchgesellschaft.
(1990) Kontrastive feministische Linguistik: Mechanismen sprachlicher Diskriminierung im
Englischen und Deutschen. München: Hueber.
(1996) with Ulrich Ammon, eds. Contrastive sociolinguistics. Berlin: Mouton de Gruyter.
Kathryn Eleri Jones, born in Chester, UK. She received her PhD in Bilingualism
from the University of Lancaster, UK. She is an Honorary Research Fellow at
the Department of Linguistics at Lancaster University, and Senior Research
Consultant for Iaith Cyf. Her main areas of research are: Bilingualism, literacy,
language and gender, language policy/planning, and language and education.
(1997) with Delyth Morris. Gender and the Welsh language. Cardiff: Equal Opportunities
Commission.
(1999) “Becoming just another alphanumeric code: Farmers’ encounters with the literacies
and discourses of agricultural bureaucracy at the livestock auction.” In Exploring
situated literacies, eds. David Barton & Mary Hamilton & Roz Ivanič. London:
Routledge, 70–90.
(2000) with Marilyn Martin Jones. Multilingual literacies: Reading and writing different
worlds. Amsterdam: Benjamins.
Gianna Marcato, born in Mirano (Venezia), Italy. She received her doctoral
degree in Classical Philology from the University of Padova. She is Professor of
Dialettologia italiana at the University of Padova. Her main areas of research
are: Italian dialectology, language variation, and empirical linguistics.
(1983) Per una metodologia della ricerca sulla lingua orale. Padova: Cleup.
(1995) ed. Donna e linguaggio. Convegno internazionale di studi. Padova: Cleup.
(2000) Isole linguistiche. Padova: Unipress.
Delyth Morris, born in Bangor, Wales, UK. She received her PhD in Sociology
from the University of Wales. She is a Lecturer in Sociology and Social Policy at
the School of Social Sciences, University of Wales at Bangor. Her main areas of
research are: Welsh language, gender and the labour market.
(1995) “Language and class fractioning in a peripheral economy.” Multilingual Matters 16:
1–15.
(1997) with Kathryn Jones. Gender and the Welsh language: A research review. Cardiff: Equal
(2001) with Glyn Williams. Language planning and language use: Welsh in a global age.
Cardiff: University of Wales Press.
Uwe Kjær Nissen, born in Schleswig, Germany. He received his PhD in

Hispanic Linguistics from the University of Southern Denmark in Odense. He
is Associate Professor of Spanish at the Institute of Language and Communica-
tion at the University of Southern Denmark. His main areas of research are:
Language and gender, Spanish syntax, translation studies, and corpus linguistics.
(1986) “Sex and gender specification in Spanish.” Journal of Pragmatics 10: 725–738.
(1997a) “Do sex-neutral and sex-specific nouns exist: The way to non-sexist Spanish?” In
Kommunikation von Geschlecht – Communication of gender, eds. Friederike Braun &
Ursula Pasero. Pfaffenweiler: Centaurus, 222–241.
(1997b) “Soziolinguistik der Syntax? Genuskongruenz im Spanischen aus feministischer
Sicht.” In Sprache und Geschlecht in der Romania. Romantistisches Kolloquium X,
eds. Wolfgang Dahmen et al. Tübingen: Gunter Narr, 321–344.
Hoa Pham, born in Danang, Vietnam. She received her PhD in Linguistics from
the University of Toronto, Canada, where she is an Instructor in Vietnamese
Language at the Department of East Asian Studies. From August 2002, she will
be Assistant Professor at the Department of African and Asian Languages and
Literatures at the University of Florida. Her main areas of research are: Viet-
namese phonology, phonetics, and language variation.
(2000) “Vietnamese learners: Markedness differential hypothesis and English consonants.” In
Proceedings of GASLA IV, University of Pittsburgh, Pennsylvania, 152–162.
(2001a) Vietnamese tone: Tone is not pitch. Ph.D. dissertation, University of Toronto.
(2001b) “A pnonetic study of Vietnamese tones: Reconsideration of the Register Flip-Flop
rule in reduplication.” In Linguistics in Potsdam, Vol. 12, Proceedings of HILP 5, eds.
Caroline Fry & Antony Dubach Green & Ruben van de Vijver. Potsdam: Universitäts-
bibliothek, 140–158.
Toril Swan, born in Sandnessjøen, Norway. She received her doctoral degree in
Historical Linguistics from the University of Tromsø. She is Professor of
English Linguistics at the Department of English at the University of Tromsø.
She has been Dean of the Faculty of Humanities since 1997. Her main areas of
reseach are: Historical linguistics, language and gender, and metaphor studies.
(1988) Sentence adverbials in English: A synchronic and diachronic investigation. Oslo: Novus.
</TARGET "not">
(1997) “From manner to subject modification: Adverbialization in English.” Nordic Journal

of Linguistics 20: 179–195.
(1999) “A feast of senses: Rhetorical devices in the prose of Salman Rushdie with special
reference to metaphors and adverbs.” Nordlit 6: 119–132.
Eva-Maria Thüne, born in Heiligenstadt, Germany. She received her doctoral

degree in Comparative Cultural Studies from the University of Osnabrück,
Germany. She is Professor of German Linguistics at the Facoltà di lingue e
letterature straniere at the University of Bologna, Italy. Her main areas of
research are: Gender studies, German as a foreign language, and German
sociolinguistics.
(1998) ed. All’inizio di tutto la lingua materna. Torino: Rosenberg & Sellier.
(1999) ed. Deutsch lehren und lernen in Italien. Padova: Unipress.
(2001) “Erinnerung auf Deutsch und Italienisch – zweisprachige Individuen erzählen.”
Muttersprache 111: 255–277.
Hong Zhang, born in Wuhan, China. She received her doctoral degree in
Anthropology from Columbia University, New York. She is Assistant Professor
of Chinese at the Department of East Asian Studies, Colby College, Waterville,
Maine. Her main areas of research are: Language and culture, gender, and
folklore.
(1991) “Liberation for whom? A review of recent western literature on women’s rights in
China.” Human Rights Tribune 2 (4): 9–12.
(1992) “‘Spare women a beating for three days, they will stand on the roof and tear the house
apart’ – Images of women in Chinese proverbs.” In Locating power. Proceedings of the
Berkeley Women and Language Conference, eds. Kira Hall et al. Berkeley: University of
California, 590–600.
<TARGET "ni" DOCINFO
AUTHOR ""
TITLE "Name index"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOFFSET "4">
Name index
A Bonkewitz, Tatjana von 203

Adriaens, Geert 100, 102 Börner-Westphal, Barbara 140
Alarcos Llorach, Emilio 272 Branch, Michael 128
Almeida, Manuel 273 Braun, Friederike 21f, 105, 112, 116, 119
Apte, Mahadev L. 134 Braunmüller, Kurt 21
Arias Barredo, Aníbal 271 Bright, William 252, 273
Arkush, R. David 79 Brouwer, Dédé 87f, 90, 97
Asher, Ronald E. 273 Brown, Roger 22
Atkinson, John 128 Bucholtz, Mary 157
Attili, Grazia 209 » 311
Bùi, Pung
Awbery, Gwenllian 315, 324f Bull, Tove 229, 240, 244
Burger, Peter 82
B Burns, Amanda M. 205
Barlow, Michael 21 Burr, Elisabeth 189, 202, 205, 214
Baron, Dennis 15 Bußmann, Hadumod 22
Bascom, William 79 Bustamente y Rivero, José Luis 262
Batliner, Anton 21 Butler, Judith 157
Bauer, Robert S. 70 Butt, John 273
Bazzanella, Carla 209
Beames, John 134 C
Beckman, Mary E. 70 Calero Fernández, María Ángeles 253, 273
Beekes, Robert S. P. 222 Callejo Gallego, Javier 275
Beito, Olav 225 Cameron, Deborah 128
Bem, Sandra Lipsitz 128 Careaga Castrillo, Pilar 275
Benedict, Paul K. 70 Casares, Julio 262
Benigni, Laura 209 Castro, Rafaela 275
Benjamin, Carmen 273 Catalá Gonzalvez, Aguas Vivas 275
Benke, Gertraud 17, 21 Ceretta, Maddalena 214
Bergvall, Victoria L. 3, 17f Chan, Anita 53
Berro-García, Adolfo 262 Chan, Marjorie K. M. 39, 53, 57f, 60f, 63f,
Bhatia, Tej K. 138 70
Bierbach, Christine 214, 261, 266, 268f, Chao, Yuen Ren 59
273, 275 Chatterji, Suniti Kumarr 134
Bing, Janet M. 3 Cheng, Chin-Chuan 49
Bjarnadóttir, Sigríður 169 Chernyshev, V. A. 134
Blakar, Rolf Mikkel 220 Cheung, Samuel Hung-nin 70f
Bloch, Jules 135 Chiang, Gregory Kuei-Ke 49, 53
Bodine, Ann 137 Chiantera, Angela 206, 208, 214
338 Name index
Claudi, Ulrike 21 Fernández de la Torre Madueño, María

Clowes, Carl 317 Dolores 253
Corbett, Greville G. 4, 7, 14f, 21, 171f, Fernández Lagunilla, Marina 274
223, 253f Fodor, Istvan 21
Cortelazzo, Manlio 193, 200, 214 Forgas Berdet, Esther 266
Craig, Colette G. 5, 21 Fornara, Orsola 209
Croll, Elisabeth 51 Francis, Nelson 21
Curzan, Anne 22 Frank, Francine W. 18, 274
Friðjónsson, Jón 166, 171, 182
D Fuertes Olivera, Pedro A. 261, 263, 273f
Daniels, Karlheinz 22 Fulkerson, David 21
Ðào, Duy Anh 282 Fung, Roxana Sukyee 66f, 70
Dardano, Maurizio 189, 202, 214
Davies, Janet 315–318 G
De Bruyne, Jacques 273 Gao, Huanian 62, 70
De Caluwe, Johan 101 Gao, Yuan 44
De Marco, Anna 209 García Meseguer, Álvaro 253, 261, 266,
De Vries, Jan W. 82 272, 274
DeFrancis, John 49, 52, 54 García Mouton, Pilar 253
Della Casa, Maurizio 214 García Pascual, Enriqueta 275
DeMello, George 274 Gardner, Neil 316
Deng, Shaojun 59 Garreta Torner, Nuria 275
Didriksen, Kirstin 176–178, 181 Gauler, Eva 21
Diosdado, Esteban 273 Gerritsen, Marinel 103
Dixon, Robert M. W. 4f, 21 Gijsbers, Inge 103
Dodd, Jan 282 Gilman, Albert 22
Donaldson, Bruce C. 105 Gold, Ann Grodzins 159
Duynstee, Rob 101 González, Iris G. 275
Gosvāmı̄, Pūranagira 141
E Gottburgsen, Anja 105
Ehrlich, Susan 19, 234 Graham, Alma 21
Einarsson, Stefán 164, 166, 171, 181 Greenberg, Joseph H. 4, 170
Elmslie, W. A. L. 79 Grierson, George Abraham 134
Emeneau, M. B. 288, 311 Griffith, T. Gwynfor 213
Engelberg, Mila 111, 113, 116f, 119f, 123, Griffiths, Bruce 329
129 Grimes, Barbara F. 57
Errazu Colás, María Ángeles 262 Grimm, Jacob 21
Ettner, Charles R. 53 Grönberg, Anna Gunnarsdotter 181
Guðmundsson, Valtýr 165, 171
F Gumperz, John J. 135
Faarlund, Jan Terje 223, 227, 247 Guru, Kamta Prasad 134
Fang, Hanquan 45
Farris, Catherine S. 47, 53 H
Ferguson, Charles A. 21 Haakana, Markku 127
Hakulinen, Auli 111, 127
Name index 339
Hall, Kira 43, 146, 157, 159f Karlsson, Fred 110, 114f, 128
Hamilton, Mykol C. 119 Kastovsky, Dieter 21
Hampares, Katherine 265, 267 Kellogg, Samuel H. 134
Härmä, Juhani 21 Khái Hþng 298f
Haugen, Einar 163f, 247 Khubchandani, Lachman M. 136, 138
Heggstad, Kolbjørn 225 Killingley, Siew-Yue 70
Heine, Bernd 21 King, Christopher R. 134
Hellinger, Marlis 18, 212, 247, 257 King, Gareth 329
Heng, J. H. 45 King, Ruth 19, 234
Hershatter, Gail 44, 48 Kiuru, Silva 110, 123
Hilmisdóttir, Helga 182 Klein, Josef 274
Hjartardóttir, Þóra Björk 176, 182 Koenen, M. J. 93
Hodge, Bob 54 Köhncke, Astrid 21
Holmes, Janet 18 Kooij, Jan G. 82, 105
Honig, Emily 44, 48 Kool-Smit, Joke E. 89f, 97
Hornscheidt, Antje 176 Köpcke, Klaus-Michael 21
Hróarsdóttir, Þorbjörg 182 Korzilius, Hubert 103
Hsu, Francis L. K. 77 Koski, Mauno 124
Huisman, Joke 100 Kowallik, Sabine 273
Hurskainen, Arvi 21 Kramarae, Cheris 10, 21
Husu, Liisa 111 Kratochvil, Paul 32f, 52
Kress, Bruno 164f, 171, 181f
I Kress, Helga 176, 182
Ibrahim, H. Muhammad 21 Kučera, Henry 21
Ingimarsdóttir, Álfheiður 182 Kuusi, Matti 125, 129
Irmen, Lisa 21 Kvaran, Guðrún 182
Kwok, Helen 58, 68
J Kyrölä, Katriina 113, 128
Jacka, Tamara 43, 70
Jain, Dhanesh Kumar 136 L
Jobin, Bettina 176 Laitinen, Lea 111
Jones, Charles 21 Lakoff, George 8, 21
Jones, Dafydd Glyn 329 Lakoff, Robin 91, 135
Jones, Kathryn 315, 324 Länsimäki, Maija 122, 126
Jónsdóttir, Margrét 164, 166, 171f Lasaosa Zazu, Asunción 271
Jónsson, Gísli 182 Lázaro Carreter, Fernando 258, 272
Jónsson, Snæbjörn 164, 166, 171f Lebsanft, Franz 262
Jussila, Raimo 126 Lehikoinen, Leila 110
Lehmann, Christian 21
K Lehmann, Winfred P. 49, 53
Kachru, Yamuna 159 Leiss, Elisabeth 21
Kähler, Gisela 268f Leiwo, Matti 127
Kam, Louie 54 Lele, Jayant K. 136
Kangas, Kea 111 Lepschy, Anna Laura 188
Karjalainen, Merja 110 Lepschy, Giulio 188
340 Name index
Levy, Marion Joseph, J. R. 77 Muraro, Luisa 204

Lewis, Mark 282
L’Hoir, Francesca Santoro 238 N
Lie, Svein 223, 227, 247 Nanda, Serena 154
Lien, Chinfa 39 Nebrija, Antonio 253
Light, Timothy 53, 61, 66, 71 Nespital, Helmut 138
Lipski, John 273 Ngô, Thanh Nhàn 283
Lledó, Eulàlià 265 à Ðình Hòa 311
Nguyen,
Löbel, Elisabeth 283 à Tài Can
Nguyen, ß 283
Lockwood, William B. 188 Nickerson, Catherine 103
López García, Ángel 267, 272f, 275 Niemelä, Pirkko 111
Lozano Domingo, Irene 267, 271, 273 Nissen, Uwe Kjær 259, 262f, 270, 273f
Lþüng, Hy V. 288 Norman, Jerry 30f, 52–54
Lyons, John 256f Nuolijärvi, Pirkko 111
Nurmi, Timo 128
M
MacKay, Donald 21 O
Madsen, Richard 53 O’Donovan, Veronica 146, 159
Magnússon, Ásgeir Blöndal 179 Oelkers, Susanne 9
Magnusson, Gunnar 176 Olivares, Carmen 262
Majapuro, Anne 22, 125 Osmond, John 315
»
Manh, Phú Tþ 304 Otero, Mercè 265
Mäntynen, Anne 127 Ouyang, Jueya 58
Marcantonio, Angela 196
Marcato, Gianna 200, 211 P
Martin, Willy 105 Pathak, Anand Swarup 141
Martín Rojo, Luisa 275 Pauwels, Anne 16–18
Martyna, Wendy 137 Pellegrini, Giovan Battista 188
Mathews, R. H. 53 Perissinotto, Giorgio 259
Matthews, Stephen 58f, 70f Pham, Hoa 21
McGregor, Ronald Stuart 134 Polak, Charlotte 101
Medina Guerra, Antonia María 253 Porto Dapena, José Álvaro 271
Mehrotra, Raja Ram 136 Pöysä, Jyrki 127
Meillet, Antoine 190 Pressini, Laura 214
Mielikäinen, Aila 111 Pretto, Anna Maria 196
Migliorini, Bruno 213 Price, Adam 316
Mills, Sara 123 Prince, Alan 39, 53
Misra, K. S. 136 Prys, Delyth 329
Mohanan, Tara 134 Pusch, Luise F. 257
Moliner, María 266, 273
Morant, Ricardo 267, 272f, 275 Q
Moreno, Montserrat 275 Qiao, Yannong 61
Moreno Marimón, Montserrat 275
Morris, Delyth 314f, 324 R
Müller, Bodo 255 Ragazzini, Guiseppe 213
Name index 341
Raheja, Gloria Goodwin 159 Smolensky, Paul 39, 53

Ramsey, S. Robert 36, 52 Snijders, Gina 102f
Rautala, Helena 115, 122f Sovijärvi, Sini 128
Ravizza, Gabriella 214 Spender, Dale 258
Renkema, Jan 92 Speroni, Sperone 208
Renzi, Lorenzo 189, 202 Spina, Stefania 206
Riolo, Salvatore 199f, 207, 214 Srivastava, Murlidhar 134
Rissel, Dorothy 273 Stacey, Judith 41, 44, 47
Rögnvaldsson, Eiríkur 170, 174f, 179, 181 Stahlberg, Dagmar 105
Romaine, Suzanne 22, 159, 227 Stefánsdóttir, Steinunn 182
Romein-Verschoor, Annie 90f, 98f Stewart, Dominic 194
Royen, Gerlach 21 Strandskogen, Åse-Berit 247
Russel, William M. 254 Strandskogen, Rolf 247
Ryen, Else 220 Strunk, Klaus 21
Suardiaz, Delia Esther 259, 261, 264, 275
S Sulkala, Helena 110
Sabatini, Alma 193f, 201–205, 210–213 Svavarsdóttir, Ásta 164, 166, 170–172,
Sabatini, Francesco 214 179, 181f
Sachdeva, Rajesh 139 Swan, Toril 231f, 236
Salvador, Gregorio 266
Sampson, Geoffrey 32 T
Samson, Amanda M. 203 Tai, James H.-Y. 44, 47
Sapir, Edward 41, 48, 52 Taillefer de Haya, Lidia 253
Scheele, Brigitte 21 Tainio, Liisa 127
Schiffman, Harold F. 134 Talbot, Mary M. 22
Schulz, Muriel R. 12, 240 Tan, Dali 47
Schutz, Rik 93 Tang, Bi 77
Schwarze, Christoph 189, 202 Tang, Ting-Chi 33f, 42, 52f
Scotton, Carol Myers 45 Tannen, Deborah 209
Scriboni, Mirella 214 Tarantino, Anthony George 49
Sczesny, Sabine 105 Terborg Schmidt, Roland 273
Secco, Gianluigi 199, 207, 214 Teschner, Richard 254
Serianni, Luca 189, 193, 196, 202, 213 Theódórsdóttir, Guðrún 168, 176, 182
Serruys, Paul 50 Thompson, Laurence C. 283, 288, 311
Seybolt, Peter J. 49, 53 Thüne, Eva-Maria 206
Shapiro, Michael C. 134 Ting, Pang-hsin 39
Sharma, Aryendra 134 Tops, Guy A. J. 105
Shetter, William Z. 105 Traustadóttir, Rannveig 176
Shih, Yu-huei 47, 53 Treichler, Paula A. 10, 21, 274
Sigmundsson, Svavar 172 Trifone, Pietro 189, 202, 213f
Sigurðardóttir, Anna 175f Tuchman, Gaye 247
Simon, Beth Lee 135, 138f
Singh, Rajendra 136 U
Sjón 168 Unger, Jonathan 53
Smith, David M. 19 Unterbeck, Barbara 21
</TARGET "ni">
342 Name index
V
Väänänen, Marjatta 115 Whitehead, Harriet 21
Vajpeyi, Kishoridass 134 Whitley, M. Stanley 273
Valentine, Tamara M. 136, 140, 145, 159 Willemyns, Roland 82
Van Alphen, Ingrid 100 Williams, Colin 315
Van Dale 93f, 97 Williams, Glyn 314, 316
Van Gessel, Han 92 Wodak, Ruth 17, 21
Van Langevelde, Ab 316 Wolf, Margery 77
Van Meurs, Frank 103 Wolfe, Susan J. 274
Van Santen, Ariane 101 Wonder, John P. 274
Vannebo, Kjell 223, 227, 247 Wong, Peggy W. Y. 70
Vargas, Ana 266 Woolard, Kathryn A. 273
Vatuk, Sylvia 136, 159 Wu, Kam-yin 58
Verbiest, Agnes 90, 97
Vesikansa, Jouko 113 Y
Vikør, Lars S. 163f, 176, 222, 247 Yip, Virginia 58f, 70f
Vilppula, Matti 122, 126 Yue-Hashimoto, Anne 70
Vincent, Nigel 188
Violi, Patrizia 189, 212 Z
Virtanen, Leea 126 Zamboni, Chiara 204, 212
» Phan 290
Vũ, Ngoc Zamora Vicente, Alonso 273
Zhu, Jiefan 79
W Zimmerman, Don H. 18
Wang, Li 61 Zingarelli, Nicola 205
Wanjin, Zhu 45 Zolli, Paolo 200
Webster, Sheila K. 22 Zubin, David 21
Wen, Hua 49
West, Candace 18
<TARGET "si" DOCINFO
AUTHOR ""
TITLE "Subject index"
SUBJECT "Impact 10"
KEYWORDS ""
SIZE HEIGHT "220"
WIDTH "150"
VOVSET "4">
Subject index
A assignment 93, 138, 140, 165–166,

Aboriginal languages 9 224–225, 318–319
abstracts see nouns arbitrary assignment 140, 190, 319
abuse, terms of 126–127, 145–146, 264, formal assignment 3, 7
311 morphological assignment 21, 138, 223
address, terms of 8, 15–16, 22, 31, 42, 43, semantic assignment 3–4, 5, 6. 7, 138,
45, 48, 49, 50, 53, 93–94, 113, 136, 140, 190, 254–255, 318, 319
145–146, 148–149, 150–152, 159, asymmetric word-pairs 12, 85, 126, 239,
202, 235–236, 261, 267–268, 270, 241–244, 260, 261, 263, 264, 300
275, 284–288, 291–311 Austroasiatic 282
see also gender switch Avadhi see Hindi
agentive see suYxes
agglutinating languages 110 B
agreement 5, 6, 7, 13, 14, 21,138–139, Bagheli see Hindi
140–143, 165, 182, 189, 194–197, Bahasa Indonesia see Indonesian
199, 223–228, 253, 254, 274, 314, Bangru see Hindi
320, 321, 322–323 Bantu 6, 21
agreement conflict 14–15, 93, 145, Basque 252, 316
167–169, 172, 180, 196–197, 258 Belizean Creole see Creole
formal agreement 3, 13, 167–169, 181, Bhojpuri see Hindi
323 biological gender see gender
semantic agreement 92, 93, 167–169, Bokmål see Norwegian
181, 227, 322–323 borrowings 86, 87, 88, 103, 110, 114, 188,
syntactic agreement 14 231, 282–283
ambiguity 116 Bosnian 6, 7
analytic languages 283 Braj see Hindi
anaphora 258 Breton 313
androcentricity see male as norm Bundeli see Hindi
animate 5, 6, 112, 138, 140, 190, 283
Arabic 5, 9, 11, 15, 17, 133, 134, 252 C
Moroccan Arabic 16, 22 Cantonese see Chinese
Arawak 252 case 20, 82, 83–84, 110, 140–142, 164, 188,
Articles 224–225 223, 283, 314
definite articles 82–83, 84, 165, Catalan 252
192–193, 202, 204, 223–224, 320, Celtic 313, 314
323 Chattisgarhi see Hindi
indefinite articles 223–224 Chinese 5, 17, 18, 22, 30–31, 57, 282–283
Cantonese Chinese 31, 52, 53, 57–72
344 Subject index
Mandarin Chinese 29–55, 57, 58–59, diminutives 84, 209, 268

60, 73–80, 133 discourse see female language, male lan-
Middle Chinese 39 guage
Standard Chinese 30–31 doing gender 3, 140, 157
Yue 31, 52, 57, 70 double gender see gender
classifiers 4–5, 6, 21, 31, 80, 283, 284, Dravidian 21, 134, 135
306–307, 310 Dutch 6, 18
clitics 197–198 Antilles Dutch 82
code-switching 136 Belgian Dutch 82, 92, 100, 101, 105
collectives see nouns Indonesian Dutch 82
common gender see gender Middle Dutch 82, 88–89
compounding 11–12, 37, 89, 111, 113, Netherlands Dutch 81–108
114, 115, 123–124, 126, 166, 173, Suriname Dutch 82
178–180, 193–194, 213, 214, 223, Dyirbal 4
230, 232, 234, 255
concord see agreement E
Confucianism 38–39, 40–41, 53, 76, 77, 79 Eastern Maroon Creole see Creole
congruence see agreement empirical studies 21, 116–122, 259–260
conversational analysis see female lan- endearment, terms of 145, 159, 296–297,
guage, male language 309
coordination 9, 13, 14–15, 21, 145, 171, English 4, 6, 7, 8, 9, 10–11, 12, 14, 15, 16,
180, 196–197, 258 17, 18, 21, 31, 53, 58, 59, 60, 82, 84,
Cornish 313 103, 112, 129, 133, 134, 137, 159,
covert gender see gender 166, 174, 176, 179, 188, 223, 225,
Creole 227, 229, 234, 236, 240, 247, 252,
Belizean Creole 4, 18 255, 265, 267, 268, 273, 275, 296,
Eastern Maroon Creole (Ndyuka) 4, 12 297, 311, 313, 314–318, 324–325,
Croatian 6, 7 328, 329
cross-addressing see gender switch American English 20, 136
Czech 5, 18 Australian English 16
epicenes see nouns
D Esperanto 274
Daco-Romanian see Romanian
Danish 5, 17, 163, 222, 255 F
declension 164 Faroese 163, 177–178, 181
default see gender female generics see generic
definiteness see articles female invisibility 10, 212, 220, 259
demonstratives see pronouns female language 17–18, 22, 57–72, 111,
derivation 11–13, 86–89, 114, 128, 127, 136–137, 147–150, 157–158,
173–174, 192–193, 194, 221, 229, 208–209, 267–269, 273, 275, 288,
231–232, 252, 255 292–293, 296–297, 298–311
derogation 3, 8, 12, 34–36, 48, 49, 50, 54, feminine generics see generic
123, 126–127, 193, 200, 208, 221, feminization 100–101, 128, 129–130,
229, 232, 237, 240, 242, 261, 262, 176–177, 180, 212–213, 221,
263, 283, 289–290 262–263, 269–272, 328
Subject index 345
Finnish 4, 6, 10, 17, 21, 22, 109–132 referential gender 6, 7, 8–9, 13, 14, 19,
Finno-Ugric 109, 110, 112 85, 91, 112, 113, 145, 165–169, 174,
Flemish see Dutch 190, 191, 192, 199, 223, 228, 230,
French 5, 6, 9, 12, 13, 15, 18, 21, 82, 101, 235, 253, 255, 322, 323
111, 112, 128, 252 social gender 6, 10–11, 14, 179, 182
Frisian 316 subgenders 7
target gender 15
G generic 7, 21, 256–260, 274
Gaelic female generics 115–116
Irish Gaelic 313 feminine generics 9, 92, 245–246, 274
Scottish Gaelic 313, 314 generic pronouns 9, 10, 14, 90, 91–93,
gender 129, 170, 171, 172, 227, 245–247,
biological gender 7 274, 288
common gender 182, 223, 224, 226, male generics 9–10, 19, 36, 111,
253–254 114–115, 116–118, 127, 128, 129,
controller gender 15 137, 201
covert gender 118–122 masculine generics 9–10, 19, 21, 91–93,
default gender 19 143–145, 169–181, 198, 201–202,
double gender 254, 263, 273–274 203, 204, 206, 210, 213, 214, 220,
gender assignment see assignment 227, 232–234, 238–239, 245–246,
gender asymmetries 123–124 255–260, 270–272, 322–323, 326
gender marking 8, 11, 119, 173, 191, German 3, 5, 6, 7, 8, 9, 10, 12, 13, 15, 16,
221, 223, 231, 232, 247, 284, 286, 18, 21, 22, 82, 84, 86–87, 101, 111,
300, 306–307, 311 112, 115, 128, 176, 188, 225, 229,
gender-neutral 7, 10, 20, 36, 92, 111, 247, 252, 254, 255, 271, 274
112, 116, 118–119, 129, 201, 234, High German see German
244 Low German 231
gender-specific 7, 9, 37, 38, 98, 100, Germanic 7, 8, 20, 82, 163, 222, 233, 238,
101, 102, 144, 166, 193–194, 206, 268
212, 233, 247, 288, 326 grammatical gender see gender
gender-suYxes see suYxes Greek 3, 6, 188
gender switch 145–146, 159 guidelines 18–19, 99–101, 127, 194, 203,
grammatical gender 4–15, 19, 20, 31, 205, 209–213, 245, 246–247, 258,
82–105, 137, 138–158, 159, 164–181, 259, 262, 265, 269–273, 275,
188, 189–199, 222–234, 252, 324–329
253–263, 265–266, 269–272, 283,
314, 318–329 H
indefinite gender 7, 8, 9–10, 11 Hebrew 6, 14, 19, 145
lexical gender 6, 7–8, 13, 14, 31, 36, 37, High German see German
84, 112–113, 166, 190–191, 224, Hindi 6, 8, 133–162
228–229, 230, 255, 275, 284, 318 Avadhi 134
loss of gender 4, 6, 7, 21 Bagheli 134
natural gender 93, 105, 137–138, 157, Bangru 134
159, 160, 166, 223 Bhojpuri 134, 135, 139
origin of gender 4, 21 Braj 134
346 Subject index
Bundeli 134 Kannada 135

Chattisgarhi 134 kinship terms 7, 8, 16, 34, 53, 113, 115,
Eastern Hindi 134, 138 136, 166, 190, 229, 230, 255, 283,
Hindustani 139 284, 285–287, 288, 291–311
Kanauji 134 Kiswahili see Swahili
Standard Hindi 134, 139, 140–145, 159
Western Hindi 134, 138, 139 L
Hindustani see Hindi language
honorifics 16 language change 18–19, 20, 52, 62,
see also address 88–89, 92–93, 94–96, 102–105, 164,
hybrid nouns see agreement conflict 175, 205–206, 212–213, 222, 223,
230–237, 241–244, 270, 271,
I 298–311
Icelandic 6, 15, 163–185 language contact 138, 328
Old Icelandic see Old Norse language critique 272, 318
idiomatic expressions 8, 16–17, 43, 45–47, language reform 16, 18–19, 43–52,
49, 50, 96–97, 114–115, 116, 53–54, 90–91, 92, 99–101, 127–128,
199–201, 238–244, 289–290 176–177, 181, 204, 209–213, 220,
inanimate 3, 14, 112, 138, 140, 190, 196, 233, 244–247, 262–263, 272,
197, 224–225, 226, 254, 258, 283, 324–329
319 Latin 3, 88, 115, 188, 238, 255
Indic lexical gender see gender
Old Indic 3 Lithuanian 3
Indo-Aryan 133, 134 loanwords see borrowings
Indo-European 5, 110, 172, 188, 189–190, Low German see German
222, 229, 313
Indonesian (Bahasa Indonesia) 5, 12, 16 M
insults see abuse male as norm (MAN) 11, 13, 19, 73, 79,
Iranian 7, 21 114–115, 118–122, 124, 128, 174,
Old Iranian 3 181, 203, 205, 208, 220
Irish see Gaelic male generics see generic
Iroquoian 9 male language 17–18, 22, 57–72,111, 127,
Italian 6, 13, 17, 187–214, 252, 258 136–137, 147–150, 157–158,
Sicilian Italian 199–201, 207–208, 214 208–209, 267–269, 273, 275, 296,
Standard Italian 188, 199–200 297, 298–311
Venetian Italian 200, 207–208 Mandarin see Chinese
J Manx 313
Japanese 5, 9, 15, 17, 18 markedness 169–171, 173, 181, 220,
Javanese 15 256–257
job advertisements 98, 100, 102–105, 176, masculine generics see generic
180, 204, 210, 323–329 matriarchy 9
job titles see occupational terms metaphors 8, 16, 115, 137, 208, 254, 264,
307–308
K Mon-Khmer 282
Kanauji see Hindi mythology 114
Subject index 347
N Old Norse 163, 223, 232, 234

Nahuatl 252 Oneida 9
names 3, 151, 152, 202, 234–236, 288, 291, Oriya 5, 6
296, 300, 301, 302, 306, 307,
308–309, 310, 311 P
family names 45, 93–94, 95–96, 106, particles
114, 202, 204, 235, 236, 248, 308 sentence-final particles 17, 57–72
first names 16, 42, 50, 235, 236 Persian 7, 21, 133, 134, 159
patronyms 248 personification 254
natural gender see gender Pidgin 134, 138, 139
Ndyuka see Creole Polish 6, 12, 18
neologism 92, 181, 205–206, 274 politeness 60, 66, 147, 150, 301, 303, 305,
neutralization 99–105, 127, 128, 176, 212, 310
221, 246, 247, 328 Prakrit 135
New Guinean languages 21 prefixes 6, 268
non-sexist language see language reform productivity 12, 87–89, 99, 100, 193, 206,
Norse see Old Norse 231, 255
Norwegian 6, 11, 12, 17, 89, 163, 219–249 professional titles see occupational terms
Bokmål 222–228, 244, 248 pronouns 9, 10–11, 31, 37, 165, 197–199,
Nynorsk 222–228, 231, 244, 248 232, 320–322, 327, 328
Riksmål 222, 223 see also generic
noun classes 5–6, 21 demonstrative pronouns 83
nouns indefinite pronouns 91, 144, 169, 170,
abstract nouns 319 171–172, 180, 199, 202, 245,
collective nouns 257 292–293
epicene nouns 144, 167, 178, 190, 225, interrogative pronouns 292–293
231, 257–258, 259–260, 270, 274, personal pronouns 7–8, 13–14, 36–37,
319 83, 112, 136, 150–151, 169, 172,
hybrid nouns see agreement conflict 197–198, 225–228, 287–288, 291,
number 83, 92, 100, 140, 142, 143, 164, 298–305, 310, 311, 320–321
188, 195, 224, 228, 252, 256, 288, possessive pronouns 83, 198, 226, 321,
320, 321–322, 326, 327 322
numeral classifiers see classifiers reflexive pronouns 225–226, 288
numerals 4, 165 relative pronouns 181, 195
Nynorsk see Norwegian prototype 2, 8
proverbs 16–17, 22, 45–47, 53, 73–80,
O 124–125, 144, 207–208, 273, 286,
obscenity 17, 264, 267 289–290, 309
occupational terms 10–13, 84–86, 90, psycholinguistic experiments see empirical
97–105, 111, 113, 115–116, 117–118, studies
124, 127, 173–175, 176, 177–180,
182, 193–194, 203, 206, 212, 230, Q
232–234, 244, 246, 253, 260, Quechua 252
262–263, 265–266, 270, 271, 274,
284, 294–295, 301, 324, 325–327
</TARGET "si">
348 Subject index
R 231–232, 255, 265, 274, 319, 326

referential gender see gender Swahili 5, 6, 21
register see honorifics Swedish 6, 110, 114, 163, 176, 182, 238
Riksmål see Norwegian synthetic languages 252
Romance 3, 7, 83, 188, 190, 251
Romanian (Daco-Romanian) 6, 11, 14, T
196 tag-questions 17, 64
Russian 3, 7, 9, 10, 12, 13, 15, 17, 19, 21 Tamil 134, 135
target see gender
S Telugu 134
Sanskrit 133, 134, 135 Tulu 135
Scandinavian 21, 222 Turkish 4, 6, 10, 11, 16–17, 21, 264
Scottish see Gaelic
self-reference 146, 148, 149, 153, 154–156, U
157, 160, 284–288, 291–311 Urdu 133, 134, 159
Semitic 5
Seneca 9 V
Serbian 6, 7 Venetian see Italian
sexist language 4, 10, 45–47, 49, 50, 51, Viet-Muong 282
90–91, 111, 127, 136, 205, 220–221 Vietnamese 5, 15, 21, 22, 281–312
Sicilian see Italian Central Vietnamese 282
Slavic 8 Northern Vietnamese 282, 296, 300,
social gender see gender 302, 303–304, 311
Spanish 6, 18, 196, 251–279 Southern Vietnamese 282, 283, 285, 287,
Castilian Spanish see Peninsular Spanish 296–297, 300, 302, 311
Iberian Spanish see Peninsular Spanish visibility strategy see feminization
Latin American Spanish 252, 263, 264,
273 W
Mexican Spanish 264, 273 Welsh 6, 12, 313–330
Peninsular Spanish 252, 264, 268, 273 word-formation see compounding, deri-
speech levels see honorifics vation
splitting 19, 92, 98, 102, 103, 104, 211, word order 15, 20, 37–39, 50, 252, 270,
213, 245, 246, 247, 259–260, 270, 283, 314
271–272, 274, 275, 326, 327 writing system 30, 32–36, 38, 48, 49–50,
stereotypes 2, 11, 20, 42, 50, 60, 63, 68–70, 52, 54, 67, 68, 69, 282
73, 96–97, 147, 203, 208, 261, 265,
266, 267, 290 Y
suYxes 268 Yue see Chinese
agentive suYxes 192–193, 231
gender-suYxes 8, 12, 86–89, 99, 100,
113–114, 128, 173–174, 190,
191–193, 206, 213, 221, 223,
In the series IMPACT: STUDIES IN LANGUAGE AND SOCIETY the following titles
have been published thus far or are scheduled for publication:
1. PÜTZ, Martin (ed.): Language Choices. Conditions, constraints, and consequences.
1997.
2. KIBBEE, Douglas A. (ed.): Language Legislation and Linguistic Rights. Selected
Proceedings of the Language Legislation and Linguistic Rights Conference, the Uni-
versity of Illinois at Urbana-Champaign, March, 1996. 1998.
3. LINELL, Per: Approaching Dialogue. Talk, interaction and contexts in dialogical
perspectives. 1998.
4. OWENS, Jonathan: Neighborhood and Ancestry. Variation in the spoken Arabic of
Maiduguri, Nigeria. 1998.
5. ANDREWS, David R.: Sociocultural Perspectives on Language Change in Diaspora.
Soviet immigrants in the United States. 1999.
6. RICENTO, Thomas (ed.): Ideology, Politics and Language Policies. Focus on English.
2000.
7. McCAFFERTY, Kevin: Ethnicity and Language Change. English in (London)Derry,
Northern Ireland. 2001.
8. ARMSTRONG, Nigel: Social and Stylistic Variation in Spoken French. A comparative
approach. 2001.
9. HELLINGER, Marlis and Hadumod BUßMANN (eds.): Gender Across Languages.
The linguistic representation of women and men: Volume I. 2001.
10. HELLINGER, Marlis and Hadumod BUßMANN (eds.): Gender Across Languages.
The linguistic representation of women and men. Volume 2. 2002.
11. In preparation.
12. OKITA, Toshie: Invisible Work. Bilingualism, language choice and childrearing in
intermarried families. 2002.
13. OAKES, Leigh: Language and National Identity. Comparing France and Sweden.
2001.
14. WEBB, Vic: Language in South Africa. The role of language in national transforma-
tion, reconstruction and development. n.y.p.
15. BOXER, Diana: Applying Sociolinguistics. Domains and face-to-face interaction.
n.y.p.

Gender Across Languages - The Linguistic Representation of Women and Men Marlis Hellinger, Hadumod Bussmann

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gender Across Languages - The Linguistic Representation of Women and Men Marlis Hellinger, Hadumod Bussmann

Uploaded by

Copyright:

Available Formats

<DOCINFO

SUBJECT "Impact 10"

SIZE HEIGHT "220"

Ulrich Ammon William Labov

John Benjamins Publishing Company

National Standard for Information Sciences – Permanence of Paper for Printed

Library of Congress Cataloging-in-Publication Data

TITLE "Table of contents"

SUBJECT "Impact 10"

SIZE HEIGHT "220"

Languages of volume I and III vii